Review of Economic Studies (2006) 73, 85–117
c 2006 The Review of Economic Studies Limited
0034-6527/05/00000000$02.00
Das Human-Kapital: A Theory of
the Demise of the Class Structure
ODED GALOR
Brown University
and
OMER MOAV
Hebrew University, Shalem Center and CEPR
First version received October 2004; final version accepted March 2005 (Eds.)
The history of society is the history of struggles between social classes.
Karl Marx
This paper suggests that the demise of the capitalists–workers class structure was a socio-economic
transformation orchestrated by the capitalists in reaction to the increasing importance of human capital in
sustaining their profit rates. Physical capital accumulation in the process of industrialisation enhanced the
importance of human capital in production and generated incentives for capitalists to support the provision
of public education for the masses, triggering the demise of the existing class structure. The implications
of the theory are consistent with the voting patterns on England’s education reform of 1902.
1. INTRODUCTION
We argue that a significant driving force in the demise of the European capitalists-workers class
structure was a socio-economic transformation triggered by the capitalists in reaction to the
increasing importance of human capital in sustaining their profit rates. The accumulation of
physical capital in the process of industrialization enhanced the importance of human capital
in production process and generated an incentive for the capitalists to support the provision of
public education for the masses, planting the seeds for the demise of the existing class structure.
Existing theories about the demise of the capitalists-workers class structure focus on the
role of a class struggle in this significant change. According to the Marxist theory, capital accumulation and the associated decline in profit rates would intensify the degree of exploitation
of workers and would bring about a class struggle between workers and capitalists that would
eventually terminate the existing class society. The recent political economy literature accepts
the basic Marxist premise regarding the significant role of the rising tension between workers
and capitalists in the changes of the class structure. It suggests, however, that social and political
reforms in Western Europe during the 19th century were an outcome of deliberate concessions
by the elite designed to avert political instability, expropriation, and possibly a revolution.
In contrast to this viewpoint, we argue that a significant force in the demise of the class
structure was a by-product of a productive cooperation between capitalists and workers. According to the theory, in the early stages of industrialization, when physical capital was the prime
engine of economic growth, societies were marked by a stable class structure characterized by a
dichotomous ownership on factors of production. Due to capital–skill complementarity, the accumulation of physical capital by the capitalists increased the importance of human capital in
85
86
REVIEW OF ECONOMIC STUDIES
sustaining the rate of return to physical capital and brought about a non-altruistic change in the
attitude of capitalists towards the provision of public education for the masses.1 The capitalists
found it beneficial to support universal publicly financed education, which enhanced the participation of the working class in the process of human and physical capital accumulation, and led to
a widening of the middle class and to the eventual demise of the capitalists–workers class structure.2 Thus, we argue that Karl Marx’s highly influential prediction about the inevitable class
struggle due to declining profit rates stemmed from an under-appreciation of the role that human
capital eventually played in the production process.
The theory is based on three central elements. First, the production process is characterized
by capital–skill complementarity.3 Capitalists therefore benefit from the aggregate accumulation
of human capital in society. Second, human capital is inherently embodied in individuals, and its
accumulation is characterized by decreasing marginal returns at the individual level. Hence, the
aggregate stock of human capital is larger if its accumulation is widely spread among individuals
in society. Capitalists therefore gain from a universal provision of education. Third, in the absence
of public education, investment in human capital is suboptimal due to borrowing constraints.
Therefore, public education enhances investment in human capital by the masses and may benefit
capitalists as well as workers.4
The theory suggests that the utilitarian support of capitalists for the provision of universal
public education was instrumental for the rapid formation of human capital and was therefore
a catalyst, and possibly even a necessary condition, for the demise of the class society. The
support for public education is unanimous among workers and capitalists who carry its prime
financial burden.5 That is, due to the coexistence of credit market imperfections and capital–skill
complementarity, the redistribution associated with public education is Pareto improving.6
The willingness of the capitalists to support universal public education rather than selective industrial education captures two of the underlying forces in the complementarity between
human capital and physical capital. First, it appears that in the second phase of the Industrial Revolution the increase in basic literacy that was associated with universal primary education raised
labour productivity. Second, investment in universal primary education generated a wider talent
pool for advanced industrial and managerial occupations, benefiting the production process at the
higher end.
1. Since firms have limited incentive to invest in the general human capital of their workers, in the presence of
credit market imperfections, the level of education would be suboptimal unless it would be financed publicly.
2. Indeed, the second phase of the Industrial Revolution was associated with a widening middle class of whitecollar workers, skilled artisans, and independent entrepreneurs (Cameron, 1989, p. 213). Moreover, the development
of the middle class was encouraged by industrialists who demanded not only a more educated labour force but also an
intermediate class of people who could serve in managerial and marketing positions (Anderson, 1975, p. 193).
3. See Goldin and Katz (1998) for evidence regarding capital–skill complementarity.
4. See Galor and Zeira (1993), Benabou (1996), Durlauf (1996), Fernandez and Rogerson (1996), and Galor and
Moav (2004) for the effect of credit market imperfections on investment in human capital and economic growth in an
unequal society. In particular, Galor and Moav (2004) offer a unified account for the effect of income inequality on the
process of development. They argue that the replacement of physical capital accumulation by human capital accumulation
as a prime engine of economic growth changed the qualitative impact of inequality on the process of development. In the
absence of an effective cooperation in the provision of public schooling, equality alleviates the adverse effect of credit
constraints and promotes human capital accumulation and economic growth.
5. The distribution of the cost of education between workers and capitalists may differ across countries due to
differences in their socio-political structure as well as their stage of development. Nevertheless, regardless of the distribution of political power in society, in light of the importance of nourishment and health for human capital formation and
labour supply, capitalists are unlikely to impose the prime financial burden on the working class as long as wages do not
significantly exceed the subsistence level of consumption.
6. This result is related to Benabou (2000), who demonstrates that when capital and insurance markets are imperfect, policies which redistribute wealth from richer to poorer individuals can have a positive net effect on aggregate output
and growth. Unlike the current study in which the support for growth-enhancing redistribution via public education is
unanimous, in Benabou (2000) redistributions are only supported by a wide consensus in a fairly homogeneous society
but face strong opposition in an unequal one. See Benabou (2002) as well.
GALOR & MOAV
DAS HUMAN-KAPITAL
87
Our thesis implies that a conflict of interest would emerge between owners of factors of
production that differ in their degree of complementarity with human capital. In particular, the
theory suggests that a conflict of interest about the timing of the implementation of growthenhancing educational policies would emerge primarily among the economic elites—industrialists
and landowners—rather than between the ruling elite and the masses.7
Historical evidence presented in Section 3 suggests that, consistent with the proposed theory,
the process of industrialization enhanced the importance of human capital in production and
induced the capitalists to lobby for the provision of universal public education. Furthermore, as
suggested by the theory, the acquisition of human capital by the working class in the second
phase of the Industrial Revolution and the associated increase in wages, in particular, relative to
the return to capital, brought about a gradual demise of the capitalist–workers class distinction.
The basic premise of this research, regarding the positive attitude of the capitalists towards
education reforms, is examined based on the voting patterns on the Balfour Act of 1902—the
proposed education reform in the U.K. that marked the consolidation of a national education system and the creation of a publicly supported secondary school system. Variations in the support
of MPs for the Balfour Act would be expected to reflect the variations in the skill intensity in the
counties they represent. Higher support for the Balfour Act would be expected from MPs who
represent industrial skill-intensive counties. The empirical analysis supports the main hypothesis.
It establishes that there exists a significant positive effect of skill intensiveness of the industrial
sector in a county on the propensity of the MPs to vote in favour of the education reform proposed
by the Balfour Act of 1902.
2. RELATED LITERATURE
The effect of social conflict on political and educational reforms was examined by Bowles and
Gintis (1975), Grossman (1994), Acemoglu and Robinson (2000), Bourguignon and Verdier
(2000), Grossman and Kim (2003), and Bertocchi and Spagat (2004), among others. They argue that reforms and redistribution from the elite to the masses diminish the tendency for sociopolitical instability and predation and may therefore stimulate investment and economic growth.
In particular, several studies examine the potential benefits for the elite from educational reforms.
Bourguignon and Verdier (2000) suggest that if political participation is determined by the education (socio-economic status) of citizens, the elite may not find it beneficial to subsidize universal
public education despite the existence of positive externalities from human capital. Grossman
and Kim (2003) argue that education decreases predation, and Bowles and Gintis (1975) suggest
that educational reforms are designed to sustain the existing social order, by displacing social
problems into the school system.
In contrast, we argue that a significant force in the demise of the class structure was a byproduct of a productive cooperation between capitalists and workers, rather than an outcome of
a divisive class struggle. Mutually beneficial reforms are also considered by Lizzeri and Persico
(2004) and Doepke and Zilibotti (2005). Lizzeri and Persico (2004) argue that provision of public
services may have served the interest of the elite as well as the masses, and therefore the extension
of franchise redirects resources from wasteful redistribution to public goods. Doepke and Zilibotti
(2005) argue that child labour regulation may benefit capitalists by inducing parents to educate
their children, increasing the average skill of the workforce. Although they place emphasis on
the political preference of the working class, since historically unions rather than factory owners
7. Galor, Moav and Vollrath (2003) examine the effect of a conflict of interest between capitalists and landowners
on education reforms. They establish theoretically and empirically the existence of a negative effect of land inequality on
public expenditure on education. Their findings support the thesis of this paper, demonstrating that even in the presence
of a class of landowners, cooperation between workers and capitalists, in the context of education reforms, may emerge.
88
REVIEW OF ECONOMIC STUDIES
were the main active campaigners for child labour regulation, the success of the unions’ action
may have been possible only because of diminished opposition from industrialists.
3. HISTORICAL EVIDENCE
Historical evidence suggests that, consistent with the proposed theory, a significant driving force
in the demise of the capitalists–workers class structure was the eagerness of the capitalists to invest
in the education of workers in reaction to the increasing importance of human capital in sustaining
their profit rates. In particular, Section 3.1 presents evidence that the process of industrialization
enhanced the importance of human capital in production and induced the capitalists to lobby
for the provision of universal public education. Section 3.2 provides evidence demonstrating
that the accumulation of human capital by the working class in the second phase of the Industrial
Revolution was associated with an increase in wages, in particular relative to the return to capital,
in line with a fading capitalists–workers class distinction. Finally, Section 3.3 presents evidence
that dispels an alternative hypothesis that political reforms during the 19th century shifted the
balance of power towards the working class and enabled workers to implement education reforms
against the will of the capitalists.
3.1. Industrial development and education reforms
Evidence suggests that the experience of the Western world throughout the various phases of the
Industrial Revolution is consistent with the hypothesis of this research about the link between
industrial development and educational reforms. The process of industrialization was characterized by a gradual increase in the relative importance of human capital for the production process
(Abramovitz, 1993). In the first phase of the Industrial Revolution, human capital had a limited
role in the production process. Education was motivated by a variety of reasons, such as religion,
social control, moral conformity, socio-political stability, social and national cohesion, military
efficiency, and the spirit of the Age of Enlightenment. The extensiveness of public education
was therefore not necessarily correlated with industrial development, and it differed across countries due to political, cultural, social, historical, and institutional factors.8 In the second phase
of the Industrial Revolution, education reforms were designed primarily to satisfy the increasing
skill requirements in the process of industrialization, reflecting the interest of capitalists in human capital formation and thus in the provision of public education.9 The evidence suggests that
in Western Europe, the economic interests of capitalists were indeed a significant driving force
behind the implementation of educational reforms.10
3.1.1. England. In the first phase of the Industrial Revolution (1760–1830), consistent with
the proposed hypothesis, capital accumulation increased significantly without a corresponding
increase in the supply of skilled labour.11 In contrast, literacy rates evolved rather slowly, and
the state devoted virtually no resources to raising the level of literacy of the masses.12 During
8. For instance, Sandberg (1979) argues that the level of human capital in Sweden prior to 1850 was larger than
the level that would have been justified by its stage of development.
9. The form of human capital that was complementary to physical capital was rather broad, including literacy,
quantitative abilities, and general knowledge, as well as work habits such as punctuality, discipline, manners, and diligence (Graff, 1987).
10. One could argue that the rise in income during the second phase of the Industrial Revolution brought about an
increase in education since people view education as a normal consumption good. The evidence that we provide in this
section clarifies that the mechanism that we underlined is a significant force behind the accumulation of human capital
during this period.
11. For instance, the investment ratio increased from 6% in 1760 to 11·7% in the year 1831 (Crafts, 1985, p. 73).
12. Cipolla (1969), Schofield (1973), and Cressy (1980) show that literacy rates have increased during the first
phase of the Industrial Revolution but at a slower rate than in the second phase.
GALOR & MOAV
DAS HUMAN-KAPITAL
89
the first stages of the Industrial Revolution, literacy was largely a cultural skill or a hierarchical symbol and had limited demand in the production process (Mitch, 1992; Mokyr, 1993,
2001). For instance, in 1841 only 4·9% of male workers and only 2·2% of female workers were
in occupations in which literacy was strictly required (Mitch, 1992, pp. 14–15). During this
period, an illiterate labour force could operate the existing technology, and economic growth was
not impeded by educational retardation.13 Workers developed skills primarily through on-the-job
training, and child labour was highly valuable (Kirby, 2003).
The development of a national public system of education in England lagged behind that
of the Continental countries by nearly half a century (Sanderson, 1995, pp. 2–10).14 Britain’s
early industrialization occurred without a direct state intervention in the development of the
minimal skills that were required in production (Green, 1990, pp. 293–294). Furthermore, as
argued by Landes (1969, p. 340) “although certain workers—supervisory and office personnel
in particular—must be able to read and do the elementary arithmetical operations in order to perform their duties, large share of the work of industry can be performed by illiterates as indeed it
was especially in the early days of the Industrial Revolution”.
England initiated a sequence of reforms in its education system since the 1830’s and literacy
rates gradually increased. The process was initially motivated by a variety of reasons such as
religion, social control, moral conformity, socio-political stability, and military efficiency, as
was the case in other European countries (e.g. Germany, France, Holland, Switzerland) that had
supported public education much earlier.15 However, in light of the modest demand for skills and
literacy by the capitalists, the level of governmental support was rather small.16
In the second phase of the Industrial Revolution, consistent with the proposed hypothesis,
the demand for skilled labour in the growing industrial sector markedly increased (Cipolla, 1969;
Kirby, 2003) and the proportion of children aged 5–14 in primary schools increased from 11% in
1855 to 25% in 1870 (Flora, Kraus and Pfenning, 1983).17 In light of the industrial competition
from other countries, capitalists started to recognize the importance of technical education for
the provision of skilled workers. As noted by Sanderson (1995, pp. 10–13), “reading . . . enabled
the efficient functioning of an urban industrial society laced with letter writing, drawing up wills,
apprenticeship indentures, passing bills of exchange, and notice and advertisement reading”.
Manufacturers argued that: “universal education is required in order to select, from the mass of
the workers, those who respond well to schooling and would make a good foreman on the shop
floor” (Simon, 1987, p. 104). Furthermore, in 1824, Alexander Galloway, the master-engineer,
reported: “I have found from the mode of managing my business, by drawings and written descriptions, a man is not of much use to me unless he can read and write. If a man applies for
work, and says he cannot read and write, he is asked no more questions” (Thompson, 1968).
13. Some have argued that the low skill requirements even declined over this period. For instance, Sanderson
(1995, p. 89) suggests that “One thus finds the interesting situation of an emerging economy creating a whole range of
new occupations which require even less literacy and education than the old ones”.
14. For instance, in his parliamentary speech in defence of his 1837 education bill, the Whig politician, Henry
Brougham, reflected upon this gap: “It cannot be doubted that some legislative effort must at length be made to remove
from this country the opprobrium of having done less for education of the people than any of the more civilised nations
on earth” (Green, 1990, pp. 10–11).
15. Wiener (1981), as well as others, argued in contrast that the educational system in Britain was initially designed
to accentuate and perpetuate class differences, however, with “little attention or status to industrial pursuits” (p. 24).
This pre-existing motivation for education had, therefore, a limited effect on the relative earnings of the elite and is thus
tangential to the hypothesis derived in this paper. Thus, although the aspects of education that were designed to accentuate
class differences may have affected the social characteristics of the elite and the masses, they have not counterbalanced
the trend towards the gradual participation of the decendants of the working class in the accumulation of physical and
human capital.
16. Even in 1869, the government funded only one-third of school expenditure (Green, 1990, pp. 6–7).
17. Job advertisements, for instance, suggest that literacy became an increasingly desired characteristic for employment as of the 1850’s (Mitch, 1993, p. 292).
90
REVIEW OF ECONOMIC STUDIES
As it became apparent that skills were necessary for the creation of an industrial society, capitalists had an increasing interest in the level of education of the masses.18 Initially, the capitalists
established the factory schools in order to educate the children they employed. The Factory Acts
of 1802 and 1833 made it mandatory for some manufacturers to set up such schools. These laws
were poorly enforced, and the factory schools were not widespread, nor were they well received
(Cipolla, 1969, pp. 66–69; Cameron, 1989, pp. 216–217; Smelser, 1991). The pure laissez-faire
policy failed in developing a proper educational system, and capitalists demanded government
intervention in the provision of education. As James Kitson, a Leeds iron-master and advocate
of technical education explained to the Select Committee on Scientific Instruction (1867–1868):
“. . . the question is so extensive that individual manufacturers are not able to grapple with it, and
if they went to immense trouble to establish schools they would be doing it in order that others
may reap the benefit” (Green, 1990, p. 295).
An additional turning point in the attitude of capitalists towards public education was the
Paris Exhibition of 1867, where the limitations of English scientific and technical education became clearly evident. Unlike the 1851 exhibition in which England won most of the prizes, the
English performance in Paris was rather poor; of the 90 classes of manufacturers, Britain dominated only in 10. Lyon Playfair, who was one of the jurors, reported that “a singular accordance
of opinion prevailed that our country has shown little inventiveness and made little progress in the
peaceful arts of industry since 1862”. This lack of progress was attributed to the fact that “France,
Prussia, Austria, Belgium and Switzerland possess good systems of industrial education and that
England possesses none” (Green, 1990, p. 296).
The government established various parliamentary investigations into the relationship between science, industry, and education, that according to the proposed theory, were designed to
address the capitalists’ outcry about the necessity of universal public education. A sequence of
reports by these committees in the years 1868–1882 underlined the inadequate training for supervisors, managers, and proprietors, as well as workers (Green, 1990, pp. 297–298). In particular,
W. E. Forster, the Vice President of the committee of the Council of Education told the House of
Commons: “Upon the speedy provision of elementary education depends our industrial prosperity . . . if we leave our work-folk any longer unskilled . . . they will become overmatched in the
competition of the world” (Hurt, 1971, pp. 223–224). They proposed to organize a state inspection of elementary and secondary schools and to provide efficient education geared towards the
specific needs of its consumers. In particular, the Royal Commission on Technical Education of
1882 confirmed that England was being overtaken by the industrial superiority of Prussia, France,
and U.S. and recommended the introduction of technical and scientific education into secondary
schools.
As argued in the proposed theory, it appears that the government gradually yielded to the
pressure by capitalists and increased its contributions to elementary as well as higher education. In the 1870 Education Act (prior to the significant extension of the franchise of 1884 that
made the working class the majority in most industrial counties), the government assumed responsibility for ensuring universal elementary education, although it did not provide either free
or compulsory education at the elementary level (Green, 1990, p. 299). School enrolment of
10 year olds increased from 40% in 1870 to 100% in 1900, the literacy rate among men increased
from 65% in the first phase of the Industrial Revolution, to nearly 100% at the end of the 19th
century (Clark, 2002), and the proportion of children aged 5–14 in primary schools increased significantly in the second half of the 19th century, from 11% in 1855 to 74% in 1900 (Flora et al.,
1983). Finally, the 1902 Balfour Act marked the consolidation of a national education system
18. As hypothesized in this paper, there was a growing consensus among workers and capitalists about the virtues of
education reforms. The labour union movement was increasingly calling for a national system of non-sectarian education
(Green, 1990, p. 302).
GALOR & MOAV
DAS HUMAN-KAPITAL
91
and created state secondary schools (Ringer, 1979). Furthermore, science and its application in
technology gained prominence (Mokyr, 1990, 2002). New universities were established with a
strong emphasis on professional training in the medical, legal, engineering, and economic studies
(Sanderson, 1995, p. 47).
3.1.2. Continental Europe. The early development of public education occurred in the
western countries of Continental Europe (e.g. Prussia, France, Sweden, and the Netherlands) well
before the Industrial Revolution. The process was motivated by a variety of reasons, such as religion, social control, moral conformity, socio-political stability, social and national cohesion, and
military efficiency (Scott, 1977; Graff, 1987). As was the case in England, massive educational
reforms occurred in the second half of the 19th century due to the rising demand for skills in the
process of industrialization (Cipolla, 1969). Technical and scientific education had been vigorously promoted as an essential element of competitiveness and economic growth (Green, 1990,
pp. 293–294).
In France, indeed, the initial development of the education system occurred well before
the Industrial Revolution, but the process was intensified and transformed to satisfy industrial
needs in the second phase of the Industrial Revolution. The early development of elementary and
secondary education in the 17th and 18th centuries was dominated by the church and religious
orders. Some state intervention in technical and vocational training was designed to reinforce development in commerce, manufacturing, and military efficiency. After the French Revolution, the
state established universal primary schools. Nevertheless, enrolment rates remained rather low.
The state concentrated on the development of secondary and higher education with the objective
of producing an effective elite to operate the military and governmental apparatus. Secondary education remained highly selective, offering general and technical instruction largely to the middle
class (Green, 1990, pp. 135–137 and 141–142). Legislative proposals during the National Convention quoted by Cubberle (1920, pp. 514–517) are revealing about the underlying motives for
education in this period: “. . . Children of all classes were to receive that first education, physical, moral and intellectual, the best adapted to develop in them republican manners, patriotism,
and the love of labour. . . They are to be taken into the fields and workshops where they may see
agricultural and mechanical operations going on. . . ”.
The process of industrialization in France and the associated increase in the demand for
skilled labour, as well as the breakdown of the traditional apprenticeship system, significantly affected the attitude towards education. State grants for primary schools were gradually increased
in the 1830’s and legislation made an attempt to provide primary education in all regions, extend
the higher education, and provide teacher training and school inspections. The number of communities without schools fell by 50% from 1837 to 1850, and as the influence of industrialists
on the structure of education intensified, education became more stratified according to occupational patterns (Anderson, 1975, pp. 15, 31). The eagerness of capitalists for rapid education reforms was reflected by the organization of industrial societies that financed schools specializing in
chemistry, design, mechanical weaving, spinning, and commerce (Anderson, 1975, pp. 86, 204).
As was the case in England, industrial competition led industrialists to lobby for the provision of public education. The Great Exhibition of 1851 and the London Exhibition of 1862
created the impression that the technological gap between France and other European nations
was narrowing and that French manufacturers ought to invest in the education of their labour
force to maintain their technological superiority. Subsequently, the reports on industrial education by commissions established in the years 1862–1865 reflected the plea of industrialists
for the provision of industrial education on a large scale and for the implementation of scientific knowledge in the industry. “The goal of modern education . . . can no longer be to form
men of letters, idle admirers of the past, but men of science, builders of the present, initiators
92
REVIEW OF ECONOMIC STUDIES
of the future”.19 (Anderson, 1975, p. 194). Education reforms in France were extensive in the
second phase of the Industrial Revolution, and by 1881 a universal, free, compulsory, and secular primary school system had been established and technical and scientific education further emphasized. Illiteracy rates among conscripts tested at the age of 20 declined gradually
from 38% in 1851–1855 to 17% in 1876–1880 (Anderson, 1975, p. 158), and the proportion
of children aged 5–14 in primary schools increased from 51% in 1850 to 86% in 1901 (Flora
et al., 1983).
In Prussia, as well, the initial steps towards compulsory education took place at the beginning of the 18th century well before the Industrial Revolution. Education was viewed at this stage
primarily as a method to unify the state (Schleunes, 1989; Tipton, 2003). In the second part of
the 18th century, education was made compulsory for all children aged 5–13. Nevertheless, these
regulations were not strictly enforced due to the lack of funding associated with the difficulty
of taxing landlords for this purpose and due to the loss of income from child labour.20 At the
beginning of the 19th century, motivated by the need for national cohesion, military efficiency,
and trained bureaucrats, the education system was further reformed, making education a secular
activity and compulsory for a three-year period, and reconstituting the Gymnasium as a state
institution providing nine years of education for the elite (Cubberley, 1920).
The process of industrialization in Prussia and the associated increase in the demand for
skilled labour led to significant pressure for educational reforms and thereby to the implementation of universal elementary schooling (Green, 1990). Taxes were imposed to finance the school
system, and teachers’ training and certification were established. Secondary schools started to
serve industrial needs as well. The Realschulen, which emphasized the teaching of mathematics and science was gradually adopted, and vocational and trade schools were founded. Total
enrolment in secondary school increased sixfold from 1870 to 1911 (Flora et al., 1983). “School
courses . . . had the function of converting the occupational requirements of public administration, commerce and industry into educational qualifications. . .” (Muller, 1987, pp. 23–24). Furthermore, the Industrial Revolution significantly affected the nature of education in German universities. German industrialists, who perceived advanced technology as the competitive edge that
could boost German industry, lobbied for reforms in the operation of universities and offered to
pay to reshape their activities so as to favour their interest in technological training and industrial
applications of basic research (McClelland, 1980, pp. 300–301).
Similarly, the structure of education in the Netherlands and Belgium reflected the interest
of capitalists in the skill formation of the masses. In the Netherlands, as early as the 1830’s, industrial schools were established and funded by private organizations, representing industrialists
and entrepreneurs. Ultimately, in the latter part of the 19th century, the state, urged by industrialists and entrepreneurs, started to support these schools (Wolthuis, 1999, pp. 92–93, 119,
139–140, 168, and 171–172). In Belgium, primary education was backward in comparison with
other European countries, and a compulsory education system was established only at the beginning of the 20th century, when the illiteracy rate was 19%. Nevertheless, industrial development
prompted the establishment of the industrial school as of 1818, financed by local industrialists.
These schools expanded rapidly in the middle of the 19th century (Mallinson, 1963).
3.1.3. United States. The process of industrialization in U.S. also increased the importance of human capital in the production process. Evidence provided by Abramovitz and David
19. L’Enseignement professionnel, ii (1864), p. 332, quoted in Anderson (1975).
20. Indicative of the realization by the elite that education was more significant for the urban production process is
the statement by Frederick II in 1779 about education: “... if they learn too much they will run off to the cities to become
secretaries or some such things” (Schleunes, 1989).
GALOR & MOAV
DAS HUMAN-KAPITAL
93
(2000) and Goldin and Katz (2001) suggests that over the period 1890–1999, the contribution of
human capital accumulation to the growth process of U.S. nearly doubled.21 As argued by Goldin
(1999), the rise of the industrial, business, and commerce sectors in the late 19th and early 20th
centuries increased the demand for managers, clerical workers, and educated sales personnel
who were trained in accounting, typing, shorthand, algebra, and commerce. Furthermore, in the
late 1910’s, technologically advanced industries demanded blue-collar craft workers who were
trained in geometry, algebra, chemistry, mechanical drawing, etc. The structure of education was
transformed in response to industrial development and the increasing importance of human capital in the production process, and high schools adapted to the needs of the modern workplace of
the early 20th century. Total enrolment in public secondary schools increased 70-fold from 1870
to 1950 (Kurian, 1994).
Nevertheless, due to differences in the structure of education finance in U.S. in comparison
to European countries, capitalists in U.S. had only limited incentives to lobby for the provision
of education and support it financially. Unlike the central role that government funding played in
the provision of public education in European countries, the evolution of the education system in
U.S. was based on local initiatives and funding. The local nature of the education initiatives
in U.S. induced community members, in urban as well as rural areas, to play a significant role
in advancing their schooling system. Capitalists, however, faced limited incentives to support the
provision of education within a county in an environment where labour was mobile across counties, and the benefits from educational expenditure in one county could be reaped by employers
in other counties.
3.2. Schooling, factor prices, and inequality
The main hypothesis of this research suggests that in the first phase of the Industrial Revolution, prior to the implementation of significant education reforms, physical capital accumulation
was the prime engine of economic growth, and the concentration of capital among the capitalist class widened wealth inequality. Once education reforms were implemented, however, the
significant increase in the return to labour relative to capital, as well as the significant increase
in the real return to labour and the associated accumulation of assets by the workers, brought
about a decline in inequality and eventually the demise of the European 19th-century class
structure.22
The theory predicts that in the first phase of the Industrial Revolution, prior to the implementation of education reforms, capital accumulation brought about a gradual increase in wages
along with an increase in the wage–rental ratio. Education reforms in the second phase of the
Industrial Revolution are expected to generate a sharp increase in real wages along with a sharp
increase in the wage–rental ratio. Finally, wealth inequality is predicted to widen in the first
phase of the Industrial Revolution and to reverse its course in the second phase, once significant
education reforms have been implemented.
Indeed, evidence from U.K. supports this hypothesis. As depicted in Figure 1(c) and (d),
based on the data-set of Clark (2002, 2005), real wages as well as the wage–rental ratio increase
21. It should be noted that literacy rates in U.S. were rather high prior to this increase in the demand for skilled
labour. Literacy rates among the white population were already 89% in 1870, 92% in 1890, and 95% in 1910 (Engerman
and Sokoloff, 2002). Education in earlier periods was motivated by social control, moral conformity, and social and
national cohesion, as well as required skills for trade and commerce.
22. The rise in inequality in mature stages of development due to technological acceleration (e.g. Galor and Tsiddon,
1997; Caselli, 1999; Galor and Moav, 2000) does not reflect a reversal in the demise of the class structure. It is not related
to class association as reflected partly by increased intergenerational mobility (e.g. Galor and Tsiddon, 1997; Maoz and
Moav, 1999; Hassler and Rodriguez-Mora, 2000).
94
REVIEW OF ECONOMIC STUDIES
Source: Flora et al. (1983), real wage (Clark, 2005), rental rates (Clark, 2002), real wage (Clark,
2005), and return to capital (Clark, 2002).
F IGURE 1
Schooling, factor prices, and inequality, England 1770–1920. The evolution of (a) the fraction of children aged 5–14 in
public primary schools: England, 1855–1920, (b) earnings inequality: England, 1820–1913, (c) wages and rental rates:
England, 1770–1920, and (d) the wage–rental ratio: England, 1770–1920
dramatically from 1870 well into the 20th century.23 These changes in factor prices reflect the
increase in enrolment rates, as depicted in Figure 1(a) (in particular the process of education
reforms from 1830 to 1870 and its consolidation in the Education Act of 1870) and its delayed
effect on the skill level per worker.24 Thus, it appears that the demise of the class structure is
indeed associated with the significant changes that occurred around 1870 in the relative returns
to the main factors of production possessed by capitalists and workers.25 As documented in the
controversial study by Williamson (1985) about the evolution of inequality in the time period
1823–1915, wealth inequality in U.K. reached a peak around 1870 and declined thereafter (Figure
1(b)), in close association with the patterns of enrolment rates and factor prices.26
23. Clark (2005) constructs three series for wages in England over this period: Farm wage, Helper wage, and
Craftsmen wage. Figures 1(c) and (d) are based on Helper wage. (Craftsmen wage generate similar time path.) Farm
wage appears less relevant given the focus of the paper. Moreover, it should be noted that the return to capital increased
moderately over this period, despite the increase in the supply of capital, reflecting technological progress, population
growth, and accumulation of human capital.
24. It should be noted that although the demand for skill labour increased in the process, the skill premium is rather
stable over this period (Clark, 2005). The lack of clear evidence about the increase in the return to human capital over
this period is not an indication for the absence of a significant increase in the demand for human capital. However, the
significant increase in schooling that took place in the 19th century, and in particular the introduction of public education
that lowered the cost of education, generated a significant increase in the supply of educated workers and operated towards
a reduction in the return to human capital.
25. Throughout the period 1873–1913 in which real wages increase significantly, the growth rate of output per
capita is explained entirely by the contributions of physical and human capital accumulation. (Matthews, Feinstein and
Odling-Smee, 1982). An increase in labour-augmenting technological change is therefore not a viable explanation for
relative and absolute increases in real wages and the decline in inequality in U.K. over this period.
26. Feinstein (1988), criticizes Williamson’s methodology of data construction, but does not provide an alternative
that refutes the hump-shaped evolution of inequality over this period.
GALOR & MOAV
DAS HUMAN-KAPITAL
95
Source: Flora et al. (1983), Morrisson and Snyder (2000), Levy-Leboyer and Bourguignon
(1990), and Levy-Leboyer and Bourguignon (1990).
F IGURE 2
Schooling, factor prices, and inequality, France 1770–1930. The evolution of (a) the fraction of children aged 5–14
in primary schools: France, 1830–1910, (b) wealth inequality: France, 1788–1929, (c) the real wage and rental rate:
France, 1820–1913, and (d) the wage–rental ratio: France, 1820–1913
Similar patterns of the effect of education on factor prices and therefore on inequality are
observed in France as well. As argued by Morrisson and Snyder (2000), wealth inequality in
France increased during the first half of the 19th century, and as depicted in Figure 2(b), started to
decline in the last decades of the 19th century in close association with the patterns of enrolment
rates and factor prices, depicted in Figure 2(a), (c), and (d). The decline in inequality in France
appears to be associated with the significant changes in the relative returns to the main factors
of production possessed by capitalists and workers in the second part of the 19th century. As
depicted in Figure 2(c) and (d), based on the data presented in Levy-Leboyer and Bourguignon
(1990), real wages as well as the wage–rental ratio increase significantly as of 1860, reflecting
the effect of the increase in enrolment rates on the skill level per worker.
The German experience is consistent with this pattern as well. Inequality in Germany started
to decline towards the end of the 19th century (Morrisson and Snyder, 2000) in association with a
significant increase in the real wages and in the wage–rental ratio from the 1880’s (Spree, 1977;
Berghahn, 1994), which is in turn related to the provision of industrial education in the second
half of the 19th century.
The link between the expansion of education and the reduction in inequality is present in
U.S. as well. Wealth inequality in U.S., which increased gradually from colonial times until the
second half of the 19th century, reversed its course at the turn of the century and maintained its declining pattern during the first half of the 20th century (Lindert and Williamson, 1976). As argued
by Goldin (2001), the emergence of the “new economy” in the early 20th century increased the
demand for educated workers. The creation of publicly funded mass modern secondary schools
from 1910 to 1940 provided general and practical education, contributed to workers’ productivity, and opened the gates for college education. This expansion facilitated social and geographic
mobility and generated a large decrease in inequality in economic outcomes.
96
REVIEW OF ECONOMIC STUDIES
Source: Flora et al. (1983).
F IGURE 3
The evolution of voting rights and school enrolment: (a) England, 1830–1925 and (b) France, 1820–1925
3.3. The timing of educational and political reforms
This research suggests that education reforms were initiated by the capitalists in reaction to the
increasing importance of human capital in sustaining their profit rates. An alternative hypothesis
may be that political reforms during the 19th century shifted the balance of power towards the
working class and enabled workers to implement education reforms against the will of the elite.27
The evidence, however, does not support this alternative hypothesis.
Education reforms took place in autocratic states that did not relinquish political power
throughout the 19th century, and major reforms occurred in societies in the midst of the process of democratization well before the stage at which the working class constituted the majority
among the voters. In particular, the most significant education reforms in U.K. were completed
before the voting majority shifted to the working class. The patterns of education and political reforms in U.K. during the 19th century are depicted in Figure 3(a). The Reform Act of
1832 nearly doubled the total electorate, but nevertheless only 13% of the voting age population was enfranchised. The artisans, the working classes, and some sections of the lower middle
classes remained outside of the political system. The franchise was extended further in the Reform Acts of 1867 and 1884 and the total electorate nearly doubled in each of these episodes.
However, working-class voters did not become the majority in all urban counties until 1884
(Craig, 1989).
The onset of England’s education reforms, and in particular, the fundamental Education Act
of 1870 and its major extension in 1880, occurred prior to the political reforms of 1884 that made
the working class the majority in most counties. As depicted in Figure 3(a), a trend of significant
increase in primary education was established well before the extension of the franchise in the
context of the 1867 and 1884 Reform Acts. In particular, the proportion of children aged 5–14
in primary schools increased fivefold (and surpassed 50%) over the three decades prior to the
qualitative extension of the franchise in 1884 in which the working class was granted a majority
in all urban counties. Furthermore, the political reforms do not appear to have any effect on the
pattern of education reform. In fact, the average growth rate of education attendance from decade
27. See, for instance, Acemoglu and Robinson (2000), where the extension of the franchise during the 19th century
is viewed as a commitment devise to ensure future income redistribution from the elite to the masses.
GALOR & MOAV
DAS HUMAN-KAPITAL
97
to decade over the period 1855–1920 reaches a peak at around the Reform Act of 1884 and starts
declining thereafter.28
A similar pattern was seen in other European countries. In France, the expanding pattern of
education preceded the major political reform that gave the voting majority to the working class.
The patterns of education and political reforms in France during the 19th century are depicted
in Figure 3(b). Prior to 1848, restrictions limited the electorate to less than 2·5% of the voting
age population. The 1848 revolution led to the introduction of nearly universal voting rights
for males. Nevertheless, the proportion of children aged 5–14 in primary schools doubled (and
exceeded 50%) over the two decades prior to the qualitative extension of the franchise in 1848 in
which the working class was granted a majority among voters. Furthermore, the political reforms
of 1848 do not appear to have any effect on the pattern of education expansion. In the Netherlands,
political reforms did not affect the trend of education expansion and the proportion of children
aged 5–14 in primary schools exceeded 60% well before the major political reforms of 1887 and
1897. Similarly, the trends of political and education reforms in Sweden, Italy, Norway, Prussia,
and Russia do not lend credence to the alternative hypothesis.
4. THE BASIC STRUCTURE OF THE MODEL
Consider a closed overlapping generations economy in a process of development. In every period
the economy produces a single homogeneous good that can be used for consumption and investment. The good is produced using physical capital and human capital. Output per capita grows
over time due to the accumulation of these factors of production.29 The stock of physical capital
in every period is the output produced in the preceding period net of consumption and human
capital investment, whereas the stock of human capital in every period is determined by the aggregate level of public education in the preceding period.30
4.1. Production of final output
Production occurs within a period according to a neoclassical, constant-returns-to-scale, production technology. The output produced at time t, Yt , is
Yt = F(K t , Ht ) ≡ Ht f (kt ) = AHt ktα ; kt ≡ K t /Ht ; α ∈ (0, 1),
(1)
where K t and Ht are the quantities of physical capital and human capital (measured in efficiency
units) employed in production at time t, and A is the level of technology.31 The production function, f (kt ), is therefore strictly increasing, strictly concave satisfying the neoclassical boundary
28. It is interesting to note, however, that the abolishment of education fees in nearly all elementary schools occurs
only in 1891, after the Reform Act of 1884, suggesting that the political power of the working class may have affected
the distribution of education cost across the population, but consistent with the proposed thesis, the decision to educate
the masses was taken independently of the political power of the working class.
29. Earlier growth models that focus on the role of physical and human capital in the process of development
include, for instance, Lucas (1988), Caballe and Santos (1993), and Mulligan and Sala-i-Martin (1993). These models
abstract from the analysis of income heterogeneity and credit market imperfections, and therefore, do not study the
incentives of the rich to subsidize the education of the poor.
30. The model abstracts from international factor movements. Land abundance in America has generated incentives
for outflow of labour from Europe to America, intensifying the problem of labour scarcity and preventing the use of labour
inflow (rather than investment in human capital) as a remedy for labour scarcity. In contrast, as argued by Taylor (1999)
and O’Rourke, Taylor and Williamson (1996), international capital outflow from Britain was significant during the 19th
century and hence could alleviate some of the need to invest in human capital in order to sustain the profit rates.
31. The introduction of technological progress would accelerate the rise in wages and may eventually trigger the
demise of the class structure even in the absence of education reforms. Nevertheless, educational reforms would have
a significant role in expediting the process. Moreover, it should be noted, that consistent with empirical evidence TFP
growth over the relevant period for this study is negligible and output growth is based primarily on factor accumulation,
as underlined in the proposed theory.
98
REVIEW OF ECONOMIC STUDIES
conditions that assure the existence of an interior solution to the producers’ profit-maximization
problem.
Producers operate in a perfectly competitive environment. Given the wage rate per efficiency
unit of labour, wt , and the rate of return to capital, rt , producers in period t choose the level of
employment of capital, K t , and efficiency units of labour, Ht , so as to maximise profits. That
is, {K t , Ht } = arg max [Ht f (kt ) − wt Ht − rt K t ]. The producers’ inverse demand for factors of
production is therefore
rt = f (kt ) = α Aktα−1 ≡ r (kt );
wt = f (kt ) − f (kt )kt = (1 − α)Aktα ≡ w(kt ).
(2)
4.2. Individuals
In every period a generation which consists of a continuum of individuals of measure 1 is born.
Each individual has a single parent and a single child. Individuals, within as well as across generations, are identical in their preferences and innate abilities. They may differ, however, in their
family wealth and thus, due to borrowing constraints, in their capability to finance investment in
human capital in the absence of public education.
Individuals live for two periods. In the first period of their lives individuals devote their
entire time for the acquisition of human capital. The acquired level of human capital increases if
their time investment is supplemented with capital investment in education. In the second period
of their lives, individuals supply their efficiency units of labour and allocate the resulting wage
income, along with their interest income, between consumption and transfers to their children.
An individual i born in period t (a member i of generation t) receives a parental transfer, bti ,
in the first period of life. A fraction τt ≥ 0 of this capital transfer is collected by the government
in order to finance public education, whereas a fraction 1 − τt is saved for future consumption.
Individuals devote their first period for the acquisition of human capital. Education is provided
publicly free of charge.32 The acquired level of human capital increases with the real resources
invested in public education. The number of efficiency units of labour of each member of generation t in period t + 1, h t+1 , is a strictly increasing, strictly concave function of the government
real expenditure on education per member of generation t, et .33
h t+1 = h(et ),
(3)
where h(0) = 1, h (0) = γ < ∞, and limet →∞ h (et ) = 0. The assumption that the slope of the
production function of human capital is finite at the origin along with the assumption that each
individual has a minimal level of human capital, h(0) > 0, even in the absence of a real expenditure on education, assure that under some market conditions investment in human capital is not
optimal.34
32. As will become apparent, once the level of public education is chosen, individuals have no incentive to acquire
private education. In particular, in early stages of development, when the tax rate τt equals 0, individuals do not acquire
education.
33. A more realistic formulation would link the cost of education to (teachers’) wages, which may vary in the
process of development. For instance, h t+1 = h(et /wt ) implies that the cost of education is a function of the number of
efficiency units of teachers that are used in the education of each individual i. As can be derived from Section 3.4, under
both formulations the optimal expenditure on education, et , is an increasing function of the capital–labour ratio in the
economy, and the qualitative results are therefore identical.
34. These assumptions are necessary in order to assure that in the early stage of development the sole engine of
growth is physical capital accumulation and there is no incentive to invest in human capital. It permits, therefore, a sharp
presentation of the results regarding institutional transition. The typically assumed Inada condition (i.e. γ is infinite) is
designed to simplify the exposition by avoiding a corner solution, but it is not a realistic assumption.
GALOR & MOAV
DAS HUMAN-KAPITAL
99
In the second-period life, a member i of generation t supplies the acquired efficiency units of
labour, h t+1 , at the competitive market wage, wt+1 . In addition, the individual receives the gross
i , is therefore
return on savings, (1 − τt )bti Rt+1 . The individual’s second-period income, It+1
i
It+1
= wt+1 h(et ) + (1 − τt )bti Rt+1 ,
(4)
where due to complete capital depreciation Rt+1 ≡ rt+1 ≡ R(kt+1 ).
Preferences of a member i of generation t are defined over second-period consumption,
i , and the transfer to their offspring, bi .35 They are represented by a non-homothetic, logct+1
t+1
linear utility function that generates the property that the average propensity to bequest is an
increasing function of wealth:36
i
i
u it = (1 − β) log ct+1
+ β log(θ + bt+1
),
(5)
where β ∈ (0, 1) and θ > 0.37
Hence, a member i of generation t allocates second-period income between consumption,
i , and transfers to the offspring, bi . That is,
ct+1
t+1
i
i
i
ct+1
+ bt+1
≤ It+1
.
(6)
i , and a non-negative transfer
The individual chooses the level of second-period consumption, ct+1
i
to the offspring, bt+1 , so as to maximize the utility function subject to the second-period budget
constraint (6).38
Hence the optimal transfer of a member i of generation t is
i
bt+1
i
= b(It+1
)≡
i
− θ)
β(It+1
if
i
It+1
>θ
0
if
i
It+1
≤ θ,
(7)
where θ ≡ θ(1 − β)/β.
35. For simplicity, we abstract from first-period consumption. It may be viewed as part of the consumption of the
parent.
36. This utility function represents preferences under which the saving rate is an increasing function of wealth. This
classical feature (e.g. Keynes, 1920; Lewis, 1954; Kaldor, 1957) is consistent with empirical evidence. Dynan, Skinner
and Zeldes (2004) find a strong positive relationship between personal saving rates and lifetime income in U.S. They
argue that their findings are consistent with models in which precautionary saving and bequest motives drive variations
in saving rates across income groups. Furthermore, Tomes (1981) and Menchik and David (1983) find evidence that the
marginal propensity to bequeath increases with wealth. The choice of a non-homothetic utility function is necessary to
assure that workers do not invest in physical capital prior to the establishment of public schooling. A choice of a homothetic utility function would not affect the results regarding the effect of capital skill complementarity on institutional
transition, but it would imply that the demise of the class structure would have necessarily occurred even in the absence
of education reforms. Nevertheless, even under homothetic preferences, educational reforms would have a significant
role in expediting the process.
37. This form of altruistic bequest motive (i.e. the “joy of giving”) is the common form in the recent literature on
income distribution and growth. It is supported empirically by Altonji, Hayashi and Kotlikoff (1997). Utility from after
tax transfers would reduce intergenerational transfers but would not affect the qualitative results. In particular, under
utility from net transfers equation(7) below would be
i − θ/(1 − τ
i
β(It+1
t+1 )), if It+1 > θ/(1 − τt+1 )
i
i ,τ
bt+1
.
= b(It+1
)
≡
t+1
i
≤ θ/(1 − τt+1 )
0,
if It+1
i , is necessarily non-negative due to the assumption that the offspring
38. It should be noted that the transfer, bt+1
has no income in the first period of life.
100
REVIEW OF ECONOMIC STUDIES
4.3. Physical capital, human capital, and output
This section demonstrates that the stocks of physical and human capital and therefore the level of
output are determined by the aggregate level of intergenerational transfers, the level of taxation,
and governmental expenditure on public education in the preceding period.
Let Bt denote the aggregate level of intergenerational transfers in period t. A fraction τt of
this capital transfer is collected by the government in order to finance public education, whereas
a fraction 1 − τt is saved for future consumption.39 The capital stock in period t + 1, K t+1 , is
therefore
K t+1 = (1 − τt )Bt ,
(8)
whereas the government tax revenues are τt Bt .
Since population is normalized to 1, the education expenditure per young individual in
period t, et , is
(9)
et = τt Bt ,
and the stock of human capital in period t + 1, Ht+1 , is therefore
Ht+1 = h(et ) = h(τt Bt ).
(10)
Hence, the capital–labour ratio kt+1 ≡ K t+1 /Ht+1 is
kt+1 =
(1 − τt )Bt
≡ k(τt , Bt ),
h(τt Bt )
(11)
where k(0, Bt ) = Bt , ∂k(τt , Bt )/∂τt < 0, ∂k(τt , Bt )/∂ Bt > 0, 40 and the output per worker in
period t + 1 is
yt+1 = A[(1 − τt )Bt ]α h(τt Bt )1−α ≡ y(τt , Bt ).
(12)
4.4. Optimal taxation
This section derives the optimal tax rate and therefore the optimal expenditure on education from
the viewpoint of each individual in society. It demonstrates that as long as taxation is used in
order to finance public schooling, there is a consensus in society regarding the desirable tax rate.
If the government would be engaged in direct transfers from the rich to the poor in addition to
the provision of public schooling, then a conflict would emerge between the classes regarding the
desirable tax rate. This would perhaps add some realism, but would obscure unnecessarily the
focus on the role of cooperative forces in the demise of the class structure.
Given that the indirect utility function is a strictly increasing function of the individual’s
second-period wealth, the optimal tax rate, τti , from the viewpoint of member i of generation t,
(and hence the optimal expenditure on education, et = τti Bt from the viewpoint of this individual,
i .
given Bt ) would maximize the individual’s second period wealth, It+1
τti = arg max[wt+1 h(τti Bt ) + (1 − τti )bti Rt+1 ],
(13)
where wt+1 = w(kt+1 ) and Rt+1 = R(kt+1 ).
39. As will become apparent, this linear tax structure is the simplest structure that would generate the transition
from a class society. It assures that the chosen level of taxation is independent of the structure of the political system.
That is, independent of the distribution of political power or voting rights among members of society. Furthermore,
capitalists could have not effectively forced the poor to finance their own education due to the proximity of the income
of the poor to the subsistence level of consumption and the positive effect of income of the outcome of the education
process.
40. ∂k(τt , Bt )/∂ Bt > 0 if and only if h(et ) − h (et )et > 0, which is satisfied given the strict concavity of h(et ) and
the positivity of h(0).
GALOR & MOAV
DAS HUMAN-KAPITAL
101
As follows from (13), noting (2) and (11) the optimal tax rate from the viewpoint of a
member i of generation t, τti , is given by41
w(kt+1 )h (τti Bt ) = R(kt+1 )
for
τti > 0 and
w(kt+1 )γ ≤ R(kt+1 )
for
τti = 0,
(14)
where kt+1 = k(τt , Bt ). Hence, given Bt , τti is determined independently of bti , and is therefore
identical for all i.42 That is τti = τt∗ for all i. Furthermore, there exists a unique capital–labour
ratio k, below which τti = 0. That is, R(
k) = w(
k)γ .
Lemma 1.
(i) The optimal tax rate in period t, τt∗ , is identical from the viewpoint of all members of
generation t and is uniquely determined.
τt∗
= τ (Bt )
>0
f or
Bt > k
=0
f or
k
Bt ≤ k = α/(1 − α)γ .
(ii) The optimal expenditure on public education, et = τ (Bt )Bt ≡ e(Bt ) from the viewpoint of
k.
each member of generation t is strictly increasing in Bt , for Bt > Proof. Noting (2), (11), and (14), it follows from the properties of h(τt Bt ) that τt∗ is
k and (2),
uniquely determined by Bt and e (Bt ) > 0, where as follows from the definition of k = α/(1 − α)γ . Hence, since the optimal tax rate in period t is identical from the viewpoint of each member
of generation t, it follows that under any political structure, the chosen tax rate in period t is
τt = τt∗ = τ (Bt ).
(15)
Proposition 1. The tax rate in period t, τt , is
⎧
⎨> 0 f or kt+1 > k
τt
.
⎩= 0 f or k
k
t+1 ≤ Proof. Since h(0) = 1, it follows from (11), (14), and Lemma 1 that kt+1 = Bt for Bt ≤ k
k. Thus the proposition follows. and hence for kt+1 ≤ 41. Substituting (2) and (11) into (13),
τti = arg max(1 − τti )α h(τti Bt )1−α Bt α [1 − α + αbti /Bt ].
The conditions in (14) follow from the optimization problem above, using (2).
42. The unanimous agreement on the tax rate is a result of the linear tax rate and the unit elasticity of substitution
between human and physical capital in production. Given a Cobb–Douglas production function, the shares of labour and
capital are constant and wage and capital income are therefore maximized if output is maximized. If the elasticity of
substitution would be larger than unity, then the poor would prefer higher taxes, whereas if the elasticity of substitution
is smaller than unity, then the rich would prefer higher taxes.
102
REVIEW OF ECONOMIC STUDIES
Corollary 1. The chosen level of taxation in every period maximizes output per worker in
the following period. That is,
τt = arg max yt+1 ≡ arg max y(τt , Bt ).
Proof. Maximizing y(τt , Bt ) with respect to τt yields the optimality conditions given by
(14), that is, the optimality conditions for the desired level of taxation from the viewpoint of each
individual. Hence, as long as the rate of return to human capital is lower than the rate of return on
k), the chosen level of investment in public education is
physical capital (i.e. as long as kt+1 ≤ 0—the level of investment that maximizes output per worker. Once the rate of return to human
capital equals the rate of return on physical capital (i.e. once kt+1 > k), the chosen investment in
public education is positive, and it maximizes output per worker.
4.5. The dynamical system
This section derives the properties of the dynamical system that governs the evolution of the
economy in the transition from a class society to a classless society. It demonstrates that the
evolution of the economy is fully determined by the evolution of intergenerational transfer within
classes in society.
Suppose that in period 0 the economy consists of two groups of individuals in their first
period of their lives—capitalists and workers. They are identical in their preferences and differ
only in their initial wealth. The Capitalists, denoted by R (Rich), are a fraction λ of all individuals
in society, who equally own the entire initial stock of wealth. The Workers, denoted by P (Poor),
are a fraction 1 − λ of all individuals in society, who have no ownership over the initial physical
capital stock.43 Since individuals are initially homogenous within a group, the uniqueness of the
solution to their optimization problem assures that their offspring who acquire the same level of
education and are taxed equally are homogenous as well . Hence, in every period a fraction λ of
all adults are homogenous descendants of the Capitalists, denoted by members of group R, and a
fraction 1 − λ are homogenous descendants of Workers, denoted by members of group P.
The optimization of groups P and R of generation t − 1 in period t > 0, determines the
aggregate intergenerational transfers in period t, Bt .
Bt = λbtR + (1 − λ)btP ≡ B(btR , btP ),
(16)
where bti is the intergenerational transfer of each member of group i in period t; i = P, R.
Hence, the capital–labour ratio in period t + 1, kt+1 , is fully determined by the intergenerational transfers of the two groups. As follows from (11), (15), (16), and Proposition 1,
kt+1 =
[1 − τ (Bt )]Bt
≡ κ(btR , btP ),
h[τ (Bt )Bt ]
(17)
where as follows from (2) and (14), and the properties of (11), ∂κ/∂bti > 0, i = R, P. Furthermore, κ(0, 0) = 0 (since in the absence of transfers and hence savings the capital stock in the
subsequent period is 0).
Since members of group R equally own the entire initial stock of wealth in period 0 and
members of group P have no ownership over the initial stock of wealth, it follows that b0R > 0
p
and b0 = 0.
43. As will become apparent this class distinction will dissipate over time. In particular, descendants of the working
class will ultimately own some physical capital.
GALOR & MOAV
Lemma 2.
period 0.
DAS HUMAN-KAPITAL
103
If b0R < k/λ then k1 < k, and thus there is no investment in public education in
p
Proof. Since b0 = 0, (11), (16), and Lemma 1, given the properties of (3), imply that
k1 = B0 = λb0R . Hence, it follows that k1 < k and thus, as follows from Proposition 1, τ0 = 0.
Consistently with empirical evidence about the process of development, it is therefore
assumed that
b0R < k/λ,
(A1)
namely, there is no investment in public education in the early stage of development.
The evolution of transfers within each group i = R, P, as follows from (7), is given by
i
= max{β[w(kt+1 )h(τ (Bt )Bt ) + (1 − τ (Bt ))bti R(kt+1 ) − θ], 0},
bt+1
i = R, P.
(18)
Since kt+1 = κ(btR , btP ), and Bt = B(btR , btP ), the evolution of transfers of each of the two groups
is fully determined by the evolution of transfers of both types of dynasties. Namely,
i
bt+1
= max{β[w(κ(btR , btP ))h(τ (B(btR , btP ))B(btR , btP ))
+ (1 − τ (B(btR , btP )))bti R(κ(btR , btP )) − θ ], 0}
≡ ψ i (btR , btP ),
i = R, P.
(19)
(20)
Thus, the dynamical system is uniquely determined by the joint-evolution of the intergenerational transfers of Workers, P and Capitalists, R. Hence, the evolution of the economy is given
by the sequence {btP , btR }∞
t=0 that satisfies in every period
P
= ψ P (btR , btP );
bt+1
R
bt+1
= ψ R (btR , btP ),
(21)
p
where b0 = 0 and b0R > 0.
5. THE PROCESS OF DEVELOPMENT
This section analyses the endogenous demise of the Capitalists–Workers class structure as the
economy evolves from early to mature stages of development.
5.1. Regime I: Physical capital accumulation
This early stage of development is characterized by a stable class structure. Capitalists generate a
higher rate of return from a direct investment in physical capital, rather than from supporting the
education of Workers that would complement their capital in the production process. Capitalists
therefore have no incentive to financially support the education of the Workers.
Regime I is defined as the time interval 0 ≤ t < t, where t + 1 is the first period in which
the capital–labour ratio exceeds k (i.e. t is the first period in which investment in human capital
takes place). In this early stage of development the capital–labour ratio in period t + 1, kt+1 ,
which determines the investment in public education in period t, is lower than k. As follows
from Proposition 1 and Corollary 1, the tax rate is 0, there is no public education, and both
groups of individuals acquire only basic skills. That is, Ht+1 = h(0) = 1.
104
REVIEW OF ECONOMIC STUDIES
Let ǩ be the level of the capital–labour ratio such that w(ǩ) = θ. As follows from (4), ǩ is
the critical level of the capital–labour ratio in time t + 1 below which in the absence of public
investment in education in period t individuals who do not receive transfers from their parents in
i
period t do not transfer income to their offspring in period t + 1. That is, It+1
≤ θ, and therefore
i
bt+1 = 0.
In order to assure that investment in human capital will begin in a period where the poor do
not invest in physical capital, it is assumed therefore that44
k ≤ ǩ.
(A2)
k = α/(1 − α)γ , Assumption A2 implies thereAs follows from (2), ǩ = [θ/(1 − α)A]1/α . Since fore that γ > (α α (1 − α)1−α A/θ)1/α .
Lemma 3. Under Assumptions A1 and A2, there are no intergenerational transfers among
workers (i.e. btP = 0) as long as public education is not established, that is,
btP = 0 for 1 ≤ t ≤ t.
Proof. As follows from Proposition 1, the definition of t, and Assumption A1 that
assures that t > 1, for 0 ≤ t < t, there is no investment in public education and hence h t+1 = 1.
P =
Hence, since Assumption A2 implies that kt ≤ ǩ and therefore w(kt ) ≤ θ, it follows that bt+1
P
P
P
max[β[w(kt+1 ) − θ ] , 0] = 0 if bt = 0. Since b0 = 0, it follows, therefore, that bt = 0 for 1 ≤
t ≤
t. The capital–labour ratio in period t + 1, as follows from (16), (17), Proposition 1, and
Lemma 3, is
(22)
t)
kt+1 = κ(btR , 0) = λbtR for t ∈ [0,
and the level of output per worker in period t + 1, yt+1 , as follows from (1) and (22), is45
yt+1 = A[λbtR ]α for t ∈ [0,
t).
(23)
The dynamics of output per worker. The evolution of output per Worker in Regime I
is driven in this regime by physical capital accumulation. The income of the Workers is not
sufficiently high to permit intergenerational transfers and therefore savings, and the evolution of
intergenerational transfers among Capitalists determines, therefore, the accumulation of physical
capital and thus the growth of output per worker over Regime I.
The evolution of the intergenerational transfers in the economy, as follows from (21) and
Lemma 3, are
⎫
R = ψ R (b R , 0) = max[β[w(λb R ) + b R R(λb R ) − θ], 0]
bt+1
⎬
t
t
t
t
for t ∈ [0,
t ),
(24)
⎭
P
bt+1 = 0
where b0R > 0 is given. Hence, in Regime I the dynamical system is fully determined by the
evolution of transfers across members of group R.
Hence, the evolution of the entire dynamical system in Regime I can be represented by
the evolution of output per worker. Since the aggregate income of the Capitalists (group R)
44. This assumption is designed to simplify the presentation of the results. As will become apparent, even if
Assumption A2 would be violated, the Capitalists would have an incentive to support the education of Workers.
45. Note that since the size of the population is 1, Yt+1 = yt+1 .
GALOR & MOAV
DAS HUMAN-KAPITAL
105
F IGURE 4
The evolution of output per worker in the process of development
is (λ(1 − α) + α)yt (where α is the share of capital in total output that is fully owned by the
Capitalists and λ(1 − α) is the labour share of group R), it follows from (7), (23), and (24) that
the evolution of output per worker in the time period t ∈ [0,
t ) is
yt+1 = max [A{β{[λ(1 − α) + α]yt − λθ }}α , 0] ≡ φ I (yt ), for yt ∈ [0,
y),
(25)
where y = A
kα .
In order to assure that the economy would ultimately take off from Regime I to Regime II
(i.e. in order to assure that consistently with empirical evidence the process of development is
marked by human capital accumulation) it is assumed that the technology is sufficiently productive. That is,
A > A,
(A3)
is the critical level of technology such that φ I (
where A
y) = y.46
I
Figure 4 depicts the properties of φ (yt ) over the interval yt ∈ (0,
y], as established in the
following lemma and corollary.
y); y = A
k α , such that the
Lemma 4. Under Assumptions A2 and A3, there exists y ∈ (0,
I
y] are
properties of φ (yt ) in the interval yt ∈ [0,
φ I (yt ) = 0,
f or
yt ≤ y
∂φ I (yt )/∂ yt > 0,
∂ 2 φ I (yt )/∂[yt ]2 < 0,
φ I (yt ) > yt ,
f or
y < yt ≤ y
f or
y < yt ≤ y
f or
yt = y.
Proof. As follows from (25), φ I (yt ) = 0 for yt ≤ y = λθ/(λ(1−α)+α), and ∂φ I (yt )/∂ yt >
0, and ∂ 2 φ I (yt )/∂[yt ]2 < 0 for y < yt ≤ y = A{α/[(1 − α)γ ]}α . Consistently with Assumption
= [1 + λ(1 − α)α βγ α θα −α ]/[β(α + (1 − α)λ)]. It should be noted that a sufficiently
46. As follows from (25), A
high level of A that satisfies Assumption A3 does not violate Assumption A2. An increase in A and γ α holding their
ratio unchanged, does not affect A2 and increases A relative to A.
106
REVIEW OF ECONOMIC STUDIES
A2, there exists a sufficiently small γ such that y > y.47 Furthermore, Assumption A3 assures
that φ I (yt ) > yt for yt = y Corollary 2. Under Assumptions A2 and A3, the dynamical system φ I (yt ) has two steadyy]; A locally stable steady state, y = 0, and an unstable
state equilibria in the interval yt ∈ [0,
y).
steady state, y u ∈ (y,
As follows from Corollary 2, if yt < y u then output per worker contracts over time and
the system converges to the steady-state equilibrium y = 0. If yt > y u then output per worker
y], crossing into Regime II. Hence, in order to assure that
expands over the entire interval (y u ,
the process of development takes off it is assumed that
y ),
y0 ∈ ( y u ,
(A4)
implying that b0R ∈ ([y u /Aλα ]1/α , [
y/Aλα ]1/α ) = ([y u /Aλα ]1/α ,
k/λ). Hence, Assumption A1
is a subset of Assumption A4.
Thus, in Regime I, output per capita increases monotonically over time due to the process of capital accumulation. Consistent with historical evidence, the wage–rental ratio, which
is uniquely determined by the Capital–Labour ratio, increases as well. Moreover, the accumulation of physical capital by the Capitalists in Regime I raises gradually the potential role of the
education of the Workers in sustaining the profit rates of the Capitalists. Ultimately, the Capitalists find it beneficial to support public education, and the economy enters into Regime II where
the process of development is fuelled by human capital accumulation as well as physical capital
accumulation.
5.2. Regime II: Education and the decline of the class structure
These later stages of development are characterized by the onset of the gradual demise of the
Capitalists–Workers class structure. The importance of human capital in sustaining the profits of
Capitalists increases sufficiently. The Capitalists find it beneficial to financially support public
education, and ultimately Workers, as well as Capitalists, are engaged in physical capital accumulation.
5.2.1. Stage I: The birth of public schooling. In Stage I of Regime II, the economy witnesses the birth of public education. The Capitalists invest in human capital as well as in physical
capital, whereas Workers acquire education financed by the Capitalists. The wage income, however, is not sufficiently high so as to permit physical capital accumulation by the Workers.
Stage I of Regime II is defined as the time interval t ≤ t < tˆ, where tˆ is the first time period
in which Workers (group P) are engaged in intergenerational transfers, permitting physical capital accumulation by the offspring. Although Workers acquire education financed by the Capitalists, their income level is not sufficiently high so as to permit transfer to their offspring.
The Capital–Labour ratio in period t + 1, as follows from (16), (17), Proposition 1, and
Lemma 3,
(1 − τt )λbtR
for t ∈ [
t, tˆ ).
(26)
kt+1 = κ(btR , 0) =
h(τt λbtR )
47. If γ < γ < γ where γ = [α α (1 − α)1−α A/θ ]1/α < γ = [α α (1 − α)1−α A/θ + α 1+α (1 − α)−α A/λθ ]1/α , then
Assumption A2 and y > y are satisfied simultaneously. Furthermore, as discussed in the previous footnote, Assumption
A3 and y > y are mutually consistent.
GALOR & MOAV
DAS HUMAN-KAPITAL
107
where ∂κ(btR , 0)/∂btR > 0. The level of output per worker in period t + 1, yt+1 , as follows from
(1) and (26), is
(27)
yt+1 = A[(1 − τt )λbtR ]α [h(τt λbtR )]1−α for t ∈ [
t, tˆ ).
The dynamics of output per worker. The evolution of output per worker in Stage I of
Regime II is driven by the accumulation of human capital as well as physical capital. The income
of the Workers is still not sufficiently high to permit intergenerational transfers. Intergenerational
transfers among the Capitalists are the sole source of physical capital accumulation, as well as
of governmental expenditure on public education, and they determine, therefore, the growth of
output per worker over Stage I of Regime II.
The evolution of the economy in the time interval t ≤ t < tˆ, as follows from (21), is given
by
R = ψ R (b R ; 0) = β[w(k
R
R
bt+1
t+1 )h(τt λbt ) + (1 − τt )bt R(kt+1 ) − θ]
t
for t ∈ [
t, tˆ ),
(28)
P = ψ P (b R ; 0) = 0
bt+1
t
where kt+1 = κ(btR , 0).
Hence, the evolution of the entire dynamical system in Stage I of Regime II can be represented by the evolution of output per worker. Since the aggregate income of the Capitalists
(group R) is (λ(1 − α) + α)yt , as is the case in Regime I, it follows from (7), (23), and (28) that
the evolution of output per worker in the time period t ∈ [
t, tˆ ) is,
yt+1 = A{(1 − τt )β{[λ(1 − α) + α]yt − λθ }}α {h(τt β{[λ(1 − α) + α]yt − λθ })}1−α
≡ φ II (yt )
for
yt ∈ [
y, ŷ ),
(29)
where y = A
k α , τt = arg max φ II (yt ), and ŷ = θ/(1 − α), is the critical level of the output per
worker such that the income level of individuals who do not receive transfer from their parents
(i.e. members of group P) equals θ. As follows from (4), as long as yt < ŷ, the Workers (members
of group P) do not transfer income to their offspring. Hence, since y = A{α/[(1 − α)γ ]}α , it
follows from Assumption A2 that ŷ > y.
In order to assure that the economy would ultimately take off from Stage I to Stage II within
Regime II it is assumed that the technology is sufficiently productive. That is,
A ≥ Â ≡
θ 1−α
,
α α β α (1 − α)1−α
(A5)
where as established in the following Lemma, Â is a sufficiently high level of technology such
that φ II ( ŷ ) > ŷ.48
If Assumption A5 is violated then there are two feasible scenarios. The economy may converge to a steady-state equilibrium in the interval (
y, ŷ ) with public education, where individuals
are identical in their level of human capital and in their wage income, but they differ in their
level of wealth. Alternatively, the economy may proceed, nevertheless, to a long-run steady-state
equilibrium above ŷ, where offspring of the Capitalists and the Workers are indistinguishable.
Figure 4 depicts the properties of φ II (yt ) over the interval yt ∈ [
y, ŷ], as derived in the
following Lemma 5, and Corollary 3.
Â], where,
48. Assumptions A3 and A5 imply that A ≥ max[ A,
1/α
θ 1−α β(α + (1 − α)λ) − α α β α (1 − α)1−α
A > Â ↔ γ >
.
λθα 2α β 1+α (1 − α)
108
REVIEW OF ECONOMIC STUDIES
Lemma 5.
y, ŷ] are
yt ∈ [
Under Assumptions A2, A3, and A5, the properties of φ II (yt ) in the interval
∂φ II (y t )/∂ y t > 0,
∂ 2 φ II (y t )/∂ y 2t < 0,
∂φ II ( ỹ)/∂ y t = ∂φ I ( ỹ)/∂ y t ,
φ II (y
t) >
and
yt .
Proof. ∂φ II (yt )/∂ yt > 0 and ∂ 2 φ II (yt )/∂ yt2 < 0, and ∂φ II ( ỹ)/∂ yt = ∂φ I ( ỹ)/∂ yt as follows from (29) and (25) noting that h(e) is strictly concave and τt = arg max yt+1 . Since τt =
arg max yt+1 , it follows from Lemma 4 that φ II (
y) = φ I (
y) > y. Moreover, for yt ∈ [
y, ŷ], yt+1 =
II
α
φ (yt ) > A{β{[λ(1 − α) + α]yt − λθ }} (i.e. yt+1 is larger than the output per worker for τt = 0).
α
α
y, ŷ], it follows that A{β{[λ(1−α)
Since (1−α)yt ≤ θ for yt ∈ [
+α]y
α t −λθ }} ≥ A[β(yt −θ)] .
βθ
α
Hence, noting that ŷ = θ/(1 − α), it follows that φ II (yt ) > A 1−α , and therefore φ II ( ŷ) > ŷ if
α
βθ α
θ
A 1−α
> 1−α
and thus if A > Â (which is satisfied by A5. Thus, since φ II (
y) > y, φ II ( ŷ) > ŷ
and ∂ 2 φ II (yt )/∂ yt2 < 0, it follows that φ II (yt ) > yt for all yt ∈ [
y, ŷ ].
Corollary 3. The dynamical system φ II (yt ) has no steady-state equilibria in the interval
y, ŷ].
yt ∈ [
y, ŷ] and
The dynamical system φ II (yt ) has no steady-state equilibria in the interval yt ∈ [
the transfers within each dynasty of type R expand over the entire interval crossing into Stage II.
Hence, in Stage I of Regime II, the economy witnesses the birth of public education. The
Capitalists invest in physical capital, and workers as well as the Capitalists acquire education
financed by the Capitalists. The wage income, however, is not sufficiently high so as to permit
physical capital accumulation by the Workers. During Stage I of Regime II, the accumulation of
physical and human capital increases wage income further, and ultimately the economy enters
Stage II of Regime II in which the wage income is sufficiently high so as to permit the accumulation of physical capital by the Workers.
5.2.2. Stage II: The demise of the class society. The accumulation of physical and human capital during Stage I of Regime II, increases wage income further and ultimately, the economy enters Stage II of Regime II in which the wage income is sufficiently high so as to permit
the accumulation of physical capital by the Workers.
Stage II of Regime II is defined as t ≥ tˆ. In this time interval all individuals acquire education and transfer income to their offspring.
The level of output per worker in Stage II of Regime II, exceeds ŷ and the wage income of
all members of society exceeds θ. Hence, it follows from (16) and (7), that
Bt = λbtR + (1 − λ)btP = β[yt − θ ].
(30)
The capital–labour ratio in period t + 1, as follows from (11) and (30), is therefore
kt+1 =
(1 − τt )β[yt − θ ]
h(τt β[yt − θ ])
for t ∈ [tˆ, ∞),
(31)
and the level of output per worker in period t + 1, yt+1 , as follows.
The evolution of output per worker. The evolution of output per worker in Stage II of
Regime II is driven by the accumulation of human capital as well as physical capital. The income
GALOR & MOAV
DAS HUMAN-KAPITAL
109
of the Workers is sufficiently high to permit intergenerational transfers and the Workers as well
as the Capitalists contribute to physical capital accumulation and governmental expenditure on
public education, and they determine, therefore, the growth of output per worker over Stage II of
Regime II.
The evolution of output per worker in Stage II of Regime II is independent of the distribution
of intergenerational transfers across classes, and hence the evolution of the economy can be fully
characterized by the evolution of output per worker. As follows from (1) and (31), the evolution
of output per worker over the time interval t > tˆ is
yt+1 = A[(1 − τt )β[yt − θ ]]α [h(τt β[yt − θ ])]1−α ≡ φ III (yt )
for
yt > ŷ,
(32)
where τt = arg max φ III (yt ), and therefore ∂φ III (yt )/∂ yt > 0. Furthermore, ∂φ II ( ŷ )/∂ yt <
∂φ III ( ŷ )/∂ yt , and it follows from the concavity and the boundary conditions of h(e) and the
aggregate production function that ∂ 2 φ III (yt )/∂ yt2 < 0, and lim yt →∞ ∂φ III (yt )/∂ yt = 0.
y, ŷ ]. It
Figure 4 depicts the properties of φ I (yt ), φ II (yt ), φ III (yt ) over the interval yt ∈ [
should be noted that since existing resources are optimally reallocated between investments in
human and physical capital, the function is smooth at ỹ (i.e. when pubic education is adopted).
However, there is a kink at ŷ (i.e. when workers start investing in physical capital) due to the fact
that a larger fraction of income and the marginal propensity to save increases.
Corollary 4. Under A2, A3, A4, and A5, output per worker yt increases monotonically in
Stage II of Regime II and converges to a steady-state equilibrium y > ŷ.
Proof. Follows directly from the properties of φ III (yt ).
Proposition 2. Under A2, A3, A4, and A5, the economy converges to a steady-state equilibrium in which the income gap between the offspring of the Capitalists and the Workers is
eliminated.
Proof. As follows from the properties of (9), (15), (30), (31), and Corollary 4, noting that
lim yt →∞ dφ III (yt )/dyt = 0, the economy converges to a unique steady-state vector (y, k, τ , h).
Since
i
= β[w(kt+1 )h t+1 + (1 − τt )bti R(kt+1 ) − θ] for t > tˆ ,
bt+1
i = P, R
(33)
where as follows from (14) h t+1 = h(kt+1 ) and τt = τ (kt+1 ), and therefore
i
bt+1
= ζ (bti , kt+1 ).
i
(34)
i
i
Hence, given k it follows that, in the steady state, bi = b where b = ζ (b , k), otherwise (since
∂ζ (bi , k)/∂bi ≥ 0) either (bi decreases (increases) for all i, and thus k decreases (increases)) or
(b R increases indefinitely and b P decreases to 0, and thus k increases) in contradiction to the
R
R
P
P
R P
stationarity of k. Hence, b = ζ (b , k), b = ζ (b , k), and k = κ(b , b ). Since ζ (b̄i , k̄) =
i
β[h̄w(k̄) − θ + (1 − τ )R(k̄)bt ], and since in Stage II of Regime II , h(ē)w(k̄) − θ > 0, it follows
from the existence of b̄i , that (1−τ )R(k̄) < 1, and therefore b̄i is unique for all i; b̄i = β[h̄w(k̄)−
θ ]/[1 − β(1 − τ )R(k̄)]. 110
REVIEW OF ECONOMIC STUDIES
The growth of income per capita in the two stages of Regime II is an outcome of physical
and human capital accumulation. Moreover, the capital–labour ratio increases, and hence, as
observed historically, the wage–rental ratio increases as well.49
In Stage II of Regime II the economy witnesses the demise of the class society. The descendants of Workers as well as Capitalists acquire human capital as well as physical capital, the
income gap between the classes narrows and vanishes in the long run, and class characteristics
fade.
5.3. Analysis
In Regime I, physical capital is scarce, the contribution of human capital to the production process does not justify investment in human capital, and the process of development is fuelled by
capital accumulation. The wage rate is lower than the critical level that would enable individuals
who do not own any capital to engage in intergenerational transfers (and thus savings). Workers, therefore, consume their entire wages; they are not engaged in saving, capital accumulation,
and intergenerational transfers. Their descendants, therefore, are also unable to engage in saving and intergenerational transfers and Workers are in a temporary steady-state equilibrium in
which there is no investment in either physical or human capital. In contrast, the income of the
Capitalists, who own the entire stock of capital in the economy, is sufficiently high, permitting
intergenerational transfers and capital accumulation. Intergenerational transfers among the Capitalists increase over time and the stock of physical capital in the economy therefore increases as
well. During this regime, physical capital accumulation by the Capitalists decreases the return to
physical capital, and the importance of potential human capital formation in sustaining the return
to capital increases. However, as long as the Capitalists generate a higher rate of return from
a direct investment in physical capital, rather than from supporting the education of Workers,
which would complement their capital in the production process, the qualitative structure of the
economy remain unchanged. The Workers remain in a poverty trap, the Capitalists get richer, and
the process of development is based solely on physical capital accumulation.
The accumulation of physical capital by the Capitalists in Regime I gradually raises the
potential role of the education of the Workers in sustaining the profit rates of the Capitalists.
Ultimately, the Capitalists find it beneficial to support public education, and the economy enters
Regime II where the process of development is fuelled by human as well as physical capital
accumulation.
Regime II is subdivided into two stages. In Stage I of Regime II, the economy witnesses the
birth of public education. The Capitalists invest in physical capital, and Workers, as well as the
Capitalists acquire education financed by the Capitalists. The wage income, however, is not high
enough to permit physical capital accumulation by the Workers. During Stage I of Regime II, the
accumulation of physical and human capital increases wage income further, and ultimately the
economy enters Stage II of Regime II in which the wage income is high enough to permit the
accumulation of physical capital by the Workers.
In Stage II of Regime II the economy witnesses the demise of the class society. The descendants of Workers as well as Capitalists acquire human capital as well as physical capital, the
income gap between the classes narrows, vanishing in the long run, and class characteristics fade.
49. The rise in k, despite the accumulation of human capital is a result of the concavity of the human capital
production function. Suppose not. If k would decline over time, the return to capital would rise and the return to education
would decline, bringing about a reduction in investment in education and a rise in investment in physical capital and thus
an increase in k, in contradiction to the assumed reduction in k.
GALOR & MOAV
DAS HUMAN-KAPITAL
111
6. EVIDENCE FROM THE BALFOUR ACT—U.K. 1902
The basic premise of this research, regarding the positive attitude of the capitalists towards education reforms, is examined based on the voting patterns on the Balfour Act of 1902—the proposed
education reform in U.K. that marked the consolidation of a national education system and the
creation of a publicly supported secondary school system. Although the formal model abstracts
from heterogeneity in the degree of complementarity between human and physical capital in different industries and thus does not permit for a conflict of interests among capitalists about the
timing of education reforms, the theory implies that variations in the degree of complementarity
between human capital, physical capital or land in various sectors would induce variations in the
level of support for education reforms among capitalists and landowners. In particular, variations
in the support of MPs for the Balfour Act would reflect the variations in the skill intensity in the
counties they represent. Higher support for the Balfour Act would be expected from MPs who
represent industrial skill-intensive counties.
We construct a data-set gathered from a variety of historical sources on the third (and final)
vote on the Balfour Act.50 The data, as described in detail in the data Appendix, combines the
home district and party affiliation for each MP with his voting record on the Balfour Act. In
addition, the data includes county-level data on the percentage of employment in skill-intensive
industries, income per capita, degree of urbanization, and religious affiliation.
As is apparent from Table 1, which summarizes the voting patterns on the Balfour Act
according to party affiliation, Conservatives and Unionists were predominantly supportive of the
Balfour Act while Liberals were predominantly opposed. Nevertheless, variations in the voting
patterns within each of the parties due to a significant number of abstentions is sufficient for a
significant identification.
We perform Ordered Probit regressions to examine the effect of the percentage of employment in industrial skill-intensive sectors in each MP’s county on the voting patterns on the Balfour
Act. As documented in Table 2, there exists a significant positive effect of the percentage of employment in industrial skill-intensive sectors in a county on the propensity of their MPs to vote
in favour of the education reform proposed by the Balfour Act of 1902.
Regression (1) in Table 2 examines the effect of the employment in industrial skill-intensive
sectors in each MP’s county on the voting patterns on the Balfour Act, controlling for county’s
per capita income. Consistent with the main hypothesis of this research, the regression shows a
significant positive effect of the ratio of employment in industrial skill-intensive sectors in each
MP’s county on the propensity to vote for the proposed education reform.
The proposed Balfour Act of 1902 provided a role for the Anglican Church in the provision
of education. As a result, the non-conformists objected to the Act. In Regression (2), therefore,
we control for the percentage of non-conformists in each county. The regression indeed shows a
significant negative effect of the fraction of non-conformists in the county on the support for the
Act by the MPs representing that county. Nevertheless, the effect of the ratio of employment in
industrial skill-intensive sectors on the vote remains significantly positive.
In Regression (3) we add dummy variables to control for counties in Scotland and Wales.
Scotland and Wales tended to view themselves as independent nations within U.K. This may
have led them to vote systematically for or against certain measures based solely on nationalistic grounds (e.g. being against universal education because it imposes “English” education).
Since counties in Scotland and Wales were characterized by lower income, a higher fraction
50. In previous education acts, the third and final vote was conducted by voice, and no record exists of the distribution of votes across MPs. Some records of votes on various amendments for these education bills do exist, but it is not
possible to specify clearly which of the votes were crucial, or whether an affirmative vote is actually in support of the
ultimate education bill.
112
REVIEW OF ECONOMIC STUDIES
TABLE 1
Voting patterns according to party affiliation and the
regional distribution of party affiliation
Liberal
Conservative
Unionist
Other
Total
Liberal
Conservative
Unionist
Other
Total
Yes
No
Abstain
Total
19
155
43
4
221
98
3
0
14
115
92
94
26
14
226
209
252
69
32
562
Scotland
Wales
England
Total
46
18
6
2
72
20
2
1
7
30
143
232
62
23
460
209
252
69
32
562
TABLE 2
The effect of the weight of the skill-intensive sector on the support for the Balfour Act
Exp. variable
Income per capita
% in skill-intensive
occupations
% non-conformists
(1)
Vote on
Balfour Act
(2)
Vote on
Balfour Act
(3)
Vote on
Balfour Act
(4)
Vote on
Balfour Act
(5)
Vote on
Balfour Act
−0·0011
(0·12)
5·7298**
(4·49)
−0·0030
(0·33)
4·1818**
(3·09)
−1·9109**
(3·95)
−0·0033
(0·36)
4·2554**
(3·20)
−1·4129
(1·59)
0·0215
(0·09)
−0·7289 **
(2·01)
−0·0118
(1·33)
2·6177*
(1·93)
0·7349
(0·67)
−0·1246
(0·50)
−0·9086 **
(2·01)
−1·7404**
(11·44)
0·0000
455
Oprobit
0·0000
455
Oprobit
0·0000
455
Oprobit
0·0000
455
Oprobit
−0·0120
(1·33)
2·6171*
(1·94)
0·7395
(0·67)
−0·1216
(0·46)
−0·9057 **
(1·99)
−1·7399**
(11·46)
0·0246
(0·10)
0·0000
455
Oprobit
Scotland dummy
Wales dummy
Party affiliation
% urban
Chi-square P-value
Total observations
Method
Note: Absolute values of t-ratios are given in parentheses. Vote on Balfour: 2, Yes; 1, Abstain; 0,
No. Party affiliation: 1, Liberal; 0, Conservative. Standard errors are adjusted by clustering by
county.
* Indicates significance at 10%.
** Indicates significance at 5%.
of Catholics, a smaller urban sector, and a smaller skill-intensive industrial sector, relative to
England, excluding these controls might cause omitted variable biases in the other coefficient
estimates. However, despite the incorporation of the Scotland and Wales dummies, the effect
of the ratio of employment in industrial skill-intensive sectors on the vote remains significantly
positive.
In Regression (4) we control for party affiliation. As is apparent from Table 1, there is a
strong correlation between party affiliation and voting patterns: Conservatives and Unionists
were predominantly supportive of the Balfour Act while Liberals were predominantly opposed.
GALOR & MOAV
DAS HUMAN-KAPITAL
113
TABLE 3
The effect of the weight of the skill-intensive sector on MP’s party affiliation
Exp. variable
Income per capita
% in skill-intensive
occupations
(1)
Party
affiliation
(2)
Party
affiliation
(3)
Party
affiliation
(4)
Party
affiliation
−0·0139
[−0·00548]
(1·03)
−7·2551**
[−2·8581]
(3·69)
−0·0119
[−0·00469]
(1·02)
−4·6358**
[−1·8287]
(2·31)
3·3908 **
[1·3375]
(4·66)
−0·0112
[−0·0044]
(0·97)
−4·3282 **
[−1·7091]
(2·23)
4·1208 **
[1·6272]
(2·78)
−0·3227
[−0·1239]
(0·72)
0·1406
[0·0559]
(0·28)
0·0003
455
Probit
0·0000
455
Probit
0·0000
455
Probit
−0·0009
[−0·0035]
(0·81)
−4·1988 **
[−1·6587]
(2·23)
4·2180 **
[1·6664]
(3·01)
−0·3936
[−0·1499]
(0·98)
0·0815
[0·03234]
(0·17)
−0·2887
[−0·11402]
(0·72)
0·0000
455
Probit
% non-conformists
Scotland dummy
Wales dummy
% urban
Chi-square P-value
Total observations
Method
Note: Absolute values of t-ratios are given in parentheses. Marginal effects are
reported in square brackets. Vote on Balfour: 2, Yes; 1, Abstain; 0, No. Party
affiliation: 1, Liberal; 0, Conservative. Standard errors are adjusted by clustering
by county.
*Indicates significance at 10%.
**Indicates significance at 5%.
Indeed, the regression shows that affiliation with the Liberal party has a highly significant negative effect on the support for the Balfour Act, and affiliation with the Conservative party therefore
has a highly significant positive effect on the support for the Act. Nevertheless, the effect of the
ratio of employment in industrial skill-intensive sectors on the vote remains significantly positive.
Further, as established in Table 3, the ratio of employment in industrial skill-intensive sectors has
a significantly positive effect on the support for the Conservative party, which, in turn, supports
the education act.
Finally, in Regression (5) of Table 2, we control for the percentage of the urban population
within each county. This is an attempt to separate between demand for education that stems from
skill-intensive industrial development and other sources of demand for education in an urban
environment. In particular, support for public education may reflect the desire to reduce crime,
which is prevalent in an urban environment.51 Interestingly, urbanization has no significant effect
on the support for the education act, and the effect of the weight of the skill-intensive industrial
sector remains significantly positive.
Table 3 reports the results of probit regressions of the effect of the ratio of employment in
the skill-intensive industrial sector in each MP’s county on party affiliation, controlling for the
percentage of employment in skill-intensive industries, per capita income, degree of urbanization, and religious affiliation in each county and incorporating the Scotland and Wales dummies.
51. Unfortunately, direct crime statistics at the county level are unavailable for this period.
114
REVIEW OF ECONOMIC STUDIES
It shows that the ratio of employment in the skill-intensive industrial sector in each MP’s county
has a significant positive effect on affiliation with the Conservative party.52
Hence, it is apparent from Tables 2 and 3 that the ratio of employment in the skill-intensive
industrial sector has a significant positive impact on the vote on the Balfour Act through two
channels. The first is the direct channel, holding party constant. The second is through its influence on the MP’s party itself.53
7. CONCLUDING REMARKS
We argue that capital accumulation in the process of industrialization gradually intensified the
relative scarcity of skilled labour and generated an incentive for human capital accumulation.
Due to the complementarity between physical and human capital in production, the capitalists
were among the prime beneficiaries of the accumulation of human capital by the masses. They
therefore had the incentive to support the provision of public education that improved their economic well-being, contributing significantly to the demise of the existing class structure. Thus,
in contrast to the conventional wisdom, the theory suggests that the demise of the capitalists–
workers class structure was affected significantly by the outcome of a productive cooperation
between capitalists and workers, rather than an outcome of a divisive class struggle. The basic
premise of this research, regarding the positive attitude of capitalists towards education reforms,
is supported empirically by a newly constructed data-set on the voting patterns on England’s
education reform proposed in the Balfour Act of 1902.
The political reforms that accompanied the education reforms in Europe during the 19th
century, may be viewed as either an attempt by the capitalists to broaden the coalition that supports vocational public education, against the wishes of the clergy and the landlords for whom
human capital was less complementary in production,54 or as a by-product of the educational
reforms that made political inequality harder to sustain or justify. One may argue that political
reforms during the 19th century shifted the balance of power towards the working class and enabled workers to implement education reforms against the will of the capitalists. The evidence,
however, does not support this alternative hypothesis. Education reforms took place in autocratic
states that did not relinquish political power throughout the 19th century, and major reforms
occurred in societies in the midst of the process of democratization well before the stage in
which the working class constituted the majority among the voters.
APPENDIX
Vote on the Balfour Act. The voting record of each MP on the Balfour Act is gathered from the supplement to the
British Parliamentary Papers, the Division Lists. The record specifies who voted in favour of and who voted against the
bill.55 The list of the names and home districts of each of the British MPs during the vote on the Balfour Act is collected
from the British Parliamentary Papers. Party affiliation for each MP during the vote on the Balfour Act is taken from
Who’s Who of British Parliament.56
52. Adjusting the probit results to obtain marginal effects, we find that a one percentage point increase in the ratio
of employment in the industrial sector raises the probability of belonging to the Liberal party by approximately 1·7%
points.
53. It should be noted that the voting patterns of the industrial-intensive counties on alternative bills in the same
year differed significantly. In particular, the share of employment in the industrial sector is an insignificant explanatory
variable for the 1902 vote on an income tax bill using an ordered probit with the same set of controls. This result suggests
that representatives from industrial-intensive counties do not vote uniformly on each bill.
54. Cultural differences across societies may have resulted in the failure of some societies to adopt efficient institutions (e.g. Greif, 1994). Therefore, the timing of education reforms relative to the process of development may differ
across countries.
55. Any member not listed is considered to have abstained. Of the 562 MPs, 226 abstained.
56. We used only those MPs in either the Liberal or Conservative parties, the dominant parties of the time. The only
other party of significance was the Unionist party, but membership in this party was not mutually exclusive with the other
GALOR & MOAV
DAS HUMAN-KAPITAL
115
Income per capita. Income per capita in each county is estimated for each county using income tax data. Source:
Hechter (2001).
Per cent in skill-intensive occupations. The proportion of the population in a county that is employed in skillintensive industrial occupations. Source: British Regional Employment Statistics, 1901, Lee (1979) and Hechter (2001).57
Based on the British Regional Employment Statistics, 1901, Hechter divides the total employment in each county into
four categories: Agricultural, Manufacturing, Middle Class, and Civil Servant. The Middle Class category consists of
skill-intensive manufacturing occupations (Mechanical Engineering, Instrument Engineering, Electrical Engineering,
Distributive Trades, Insurance/Banking/Finance, Professional and Scientific, and Miscellaneous). Employment in theses occupations is used in the regression to capture the percentage of employment in skill-intensive occupations.
Per cent non-conformists. The proportion of the population in a county that is non-conformists. Source: Hechter
(2001).
Per cent urban. The proportion of the population in a county that resides in urban areas. Source: Hechter (2001).
Acknowledgements. The authors are grateful for comments from two anonymous referees, the Editor, Fabrizio
Zilibotti, Daron Acemoglu, Josh Angrist, George Akerlof, Jordi Caballe, Robert Gary-Bobo, Edward Glaeser, Claudia
Goldin, Herschel Grossman, Larry Katz, Victor Lavy, Joram Mayshar, Joel Mokyr, Daniele Paserman, Jim Robinson,
David Weil and seminar participants at the NBER Summer Institute, 2000, the Euroconference on Innovation, Economic
Growth, Barcelona, 2000, DIGIT, Vienna, 2001, CEPR Conference, Copenhagen, 2001, AEA, 2002, and CEPR Conference, Paris, 2003. Paula Green, Daniel Mejia, Ruti Polachek, Tamar Roth, Analia Schlosser, and Dietrich Vollrath
provided excellent research assistance. The authors’ research was supported by the ISF. Galor’s research is supported by
NSF Grant SES-0004304.
REFERENCES
ABRAMOVITZ, M. (1993), “The Search of the Sources of Growth: Areas of Ignorance, Old and New”, Journal of
Economic History, 53, 217–243.
ABRAMOVITZ, M. and DAVID, P. A. (2000), “American Macroeconomic Growth in the Era of Knowledge-Based
Progress: The Long-Run Perspective”, in S. L. Engerman and R. E. Gallman (eds.) The Cambridge Economic
History of the United States (Cambridge and New York: Cambridge University Press) 1–92.
ACEMOGLU, D. and ROBINSON, J. A. (2000), “Why Did the West Extend the Franchise? Democracy, Inequality and
Growth in Historical Perspective”, Quarterly Journal of Economics, 115, 1167–1199.
ALTONJI, J. G., HAYASHI, F. and KOTLIKOFF, L. J. (1997), “Parental Altruism and Inter Vivos Transfers: Theory and
Evidence”, Journal of Political Economy, 105, 1121–1166.
ANDERSON, R. D. (1975) Education in France 1848–1870 (Oxford: Clarendon Press).
BENABOU, R. (1996), “Inequality and Growth”, NBER Macroeconomics Annual, 11, 11–74.
BENABOU, R. (2000), “Unequal Societies: Income Distribution and the Social Contract”, American Economic Review,
90, 96–129.
BENABOU, R. (2002), “Tax and Education Policy in a Heterogeneous-Agent Economy: What Levels of Redistribution
Maximize Growth and Efficiency?”, Econometrica, 70, 481–517.
BERGHAHN, V. R. (1994) Imperial Germany, 1871–1914, Economy, Society, Culture and Politics (Providence:
Berghahn Books).
BERTOCCHI, G. and SPAGAT, M. (2004), “The Evolution of Modern Educational Systems: Technical Vs. General
Education, Distributional Conflict and Growth”, Journal of Development Economics, 73, 559–582.
BOURGUIGNON, F. and VERDIER, T. (2000), “Oligarchy, Democracy, Inequality and Growth”, Journal of Development Economics, 62, 285–313.
BOWLES, S. and GINTIS, H. (1975), “Capitalism and Education in the United States”, Socialist Revolution, 5, 101–138.
CABALLE, J. and SANTOS, M. (1993), “On Endogenous Growth with Physical and Human Capital”, Journal of Political Economy, 101, 1042–1067.
CAMERON, R. (1989) A Concise Economic History of the World from Paleolithic Times to the Present (Oxford: Oxford
University Press).
CASELLI, F. (1999), “Technological Revolutions”, American Economic Review, 89, 78–102.
CIPOLLA, C. M. (1969) Literacy and Development in the West (Harmondsworth: Penguin Books).
CLARK, G. (2002), “Farmland Rental Values and Agrarian History: England and Wales, 1500-1912” , European Review
of Economic History, 6, 281–309.
two. Many members of the Liberal and Conservative parties were Unionists as well. In addition, there were a number of
Unionist-only members. Excluding pure Unionists and other members of smaller parties, we removed 101 MPs from the
sample. In addition, we were unable to locate party affiliation data on six of the MPs, and removed them from the sample.
57. The British Regional Employment Statistics provides a breakdown of employment by industry in each county
in U.K., an area that encompasses several districts. Unfortunately, the employment data is not at a district level and each
MP was therefore assigned the percentage appropriate to the county within which his district existed.
116
REVIEW OF ECONOMIC STUDIES
CLARK, G. (2005), “The Condition of the Working-Class in England, 1200-2000”, Journal of Political Economy, 113,
1307–1340.
CRAFTS, N. (1985) British Economic Growth During the Industrial Revolution (Oxford: Clarendon Press).
CRAIG, F. W. S. (1989) British Electoral Facts, 1832–1987 (Brookfield: Gower Press).
CRESSY, D. (1980) Literacy and the Social Order: Reading and Writing in Tudor and Stuart England (Cambridge:
Cambridge University Press).
CUBBERLEY, E. P. (1920) The History of Education (Cambridge: Cambridge University Press).
DOEPKE, M. and ZILIBOTTI, F. (2005), “The Macroeconomics of Child Labor Regulation”, American Economic
Review, 1492–1524.
DURLAUF, S. N. (1996), “A Theory of Persistent Income Inequality”, Journal of Economic Growth, 1, 75–94.
DYNAN, K. E., SKINNER, J. and ZELDES, S. (2004), “Do the Rich Save More?”, Journal of Political Economy, 112,
397–444.
ENGERMAN, S. and SOKOLOFF, K. L. (2002), “Factor Endowment, Inequality, and Paths of Development among New
World Economies” (Working Paper 2959, NBER).
FEINSTEIN, C. (1988), “The Rise and Fall of the Williamson Curve”, Journal of Economic History, 48, 699–729.
FERNANDEZ, R. and ROGERSON, R. (1996), “Income Distribution, Communities, and the Quality of Public Education”, Quarterly Journal of Economics, 111, 135–164.
FLORA, P., KRAUS, F. and PFENNING, W. (1983) State, Economy and Society in Western Europe 1815–1975, Vol. I
(Chicago: St. James Press).
GALOR, O. and MOAV, O. (2000), “Ability Biased Technological Transition, Wage Inequality and Growth”, Quarterly
Journal of Economics, 115, 469–498.
GALOR, O. and MOAV, O. (2004), “From Physical to Human Capital: Inequality in the Process of Development”,
Review of Economic Studies, 71, 1001–1026.
GALOR, O., MOAV, O. and VOLLRATH, D. (2003), “Land Inequality and the Emergence of Human Capital Promoting
Institutions” (Working Paper, Brown University).
GALOR, O. and TSIDDON, D. (1997), “Technological Progress, Mobility, and Growth”, American Economic Review,
87, 363–382.
GALOR, O. and ZEIRA, J. (1993), “Income Distribution and Macroeconomics”, Review of Economic Studies, 60, 35–52.
GOLDIN, C. (1999), “Egalitarianism and the Returns to Education During the Great Transformation of American
Education”, Journal of Political Economy, 107, S65–S94.
GOLDIN, C. (2001), “The Human Capital Century and American Leadership: Virtues of the past”, Journal of Economic
History, 61, 263–292.
GOLDIN, C. and KATZ, L. F. (1998), “The Origins of Technology-Skill Complementary”, Quarterly Journal of
Economics, 113, 693–732.
GOLDIN, C. and KATZ, L. F. (2001), “The Legacy of U.S. Educational Leadership: Notes on Distribution and Economic
Growth in the 20th Century”, American Economic Review, 91, 18–23.
GRAFF, H. J. (1987) The Labyrinths of Literacy: Reflections on Literacy Past and Present (London: Falmer Press).
GREEN, A. (1990) Education and State Formation (New York: St. Martin’s Press).
GREIF, A. (1994), “Cultural Beliefs and the Organization of Society—A Historical and Theoretical Reflection on
Collectivist and Individualist Societies”, Journal of Political Economy, 102, 912–950.
GROSSMAN, H. I. (1994), “Production, Appropriation, and Land-Reform”, American Economic Review, 84, 705–712.
GROSSMAN, H. I. and KIM, M. (2003), “Education Policy: Egalitarian or Elitist?”, Economics and Politics, 15,
225–246.
HASSLER, J. and RODRIGUEZ-MORA, J. (2000), “Intelligence, Social Mobility, and Growth”, American Economic
Review, 90, 888–908.
HECHTER, M. (2001) Internal Colonialism Study: National Integration in the British Isles, 1851–1966 (Ann Arbor:
Inter-University Consortium for Political and Social Research).
HURT, J. (1971) Education in Evolution (London: Paladin).
KALDOR, N. (1957), “A Model of Economic Growth”, Economic Journal, 67, 591–624.
KEYNES, J. M. (1920) The Economic Consequences of the Peace (London: Macmillan).
KIRBY, P. (2003) Child Labor in Britain, 1750–1870 (Basingstoke: Palgrave-Macmillan).
KURIAN, G. T. (1994) Datapedia of U.S. 1790–2000, America Year by Year (Lonham, MD: Bernan Press).
LANDES, D. S. (1969) The Unbound Prometheus. Technological Change and Industrial Development in Western Europe
from 1750 to the Present (Cambridge: Cambridge University Press).
LEE, C. (1979), British Regional Employment Statistics, 1841–1971 (Cambridge: Cambridge University Press).
LEVY-LEBOYER, M. and BOURGUIGNON, F. (1990) The French Economy in the Nineteenth Century (Cambridge:
Cambridge University Press).
LEWIS, W. A. (1954), “Economic Development with Unlimited Supply of Labor”, The Manchester School, 22,
139–191.
LINDERT, P. H. and WILLIAMSON, J. G. (1976), “Three Centuries of American Inequality”, Research in Economic
History, 1, 69–123.
LIZZERI, A. and PERSICO, N. (2004), “Why Did the Elites Extend the Suffrage? Democracy and the Scope of Government, With an Application to Britain’s Age of Reform”, Quarterly Journal of Economics, 119, 707–765.
LUCAS, R. E. (1988), “On the Mechanics of Economic Development”, Journal of Monetary Economics, 22, 3–42.
MALLINSON, V. (1963) Power & Politics in Belgian Education 1815 to 1961 (London: Heinemann Educational Books).
GALOR & MOAV
DAS HUMAN-KAPITAL
117
MAOZ, Y. D. and MOAV, O. (1999), “Intergenerational Mobility and the Process of Development”, Economic Journal,
109, 677–697.
MATTHEWS, R. C., FEINSTEIN, C. H. and ODLING-SMEE, J. C. (1982) British Economic Growth 1856–1973
(Stanford: Stanford University Press).
McCLELLAND, C. E. (1980) State, Society, and University in Germany: 1700–1914 (Cambridge: Cambridge University
Press).
MENCHIK, P. and DAVID, M. (1983), “Income Distribution, Lifetime Savings, and Bequests”, American Economic
Review, 73, 672–690.
MITCH, D. (1992) The Rise of Popular Literacy in Victorian England: The Influence of Private Choice and Public Policy
(Philadelphia: University of Pennsylvania Press).
MITCH, D. (1993), “The Role of Human Capital in the First Industrial Revolution”, in J. Mokyr (ed.) The British
Industrial Revolution: An Economic Perspective (Boulder: Westview Press) 267–307.
MOKYR, J. (1990) The Lever of Riches (New York: Oxford University Press).
MOKYR, J. (1993), “The New Economic History and the Industrial Revolution”, in J. Mokyr (ed.) The British Industrial
Revolution: An Economic Perspective (Boulder, CO: Westview Press) 1–131.
MOKYR, J. (2001), “The Rise and Fall of the Factory System: Technology, Firms, and Households since the Industrial
Revolution”, Carnegie-Rochester Conference Series on Public Policy, 55, 1–45.
MOKYR, J. (2002) The Gifts of Athena: Historical Origins of the Knowledge Economy (Princeton: Princeton University
Press).
MORRISSON, C. and SNYDER, W. (2000), “Income Inequalities in France from the Early Eighteenth Century to 1985”,
Revue Economique, 51, 119–154.
MULLER, D. K. (1987), “The Process of Systematization: The Case of German Secondary Education”, in D. K. Muller,
F. Ringer and B. Simon (eds.) The Rise of the Modern Educational System (Cambridge: Cambridge University
Press) 15–52.
MULLIGAN, C. B. and SALA-I-MARTIN, X. (1993), “Transitional Dynamics in Two Sector Models of Endogenous
Growth”, Quarterly Journal of Economics, 108, 739–773.
O’ROURKE, K. H., TAYLOR, A. M. and WILLIAMSON, J. G. (1996), “Factor Price Convergence in the Late Nineteenth
Century”, International Economic Review, 37, 499–530.
RINGER, F. (1979), Education and Society in Modern Europe (Bloomington: Indiana University Press).
SANDBERG, L. G. (1979), “The Case of the Impoverished Sophisticate: Human Capital and Swedish Economic Growth
before World War I”, The Journal of Economic History, 39, 225–241.
SANDERSON, M. (1995) Education, Economic Change and Society in England 1780–1870 (Cambridge: Cambridge
University Press).
SCHLEUNES, K. A. (1989) Schooling and Society: The Politics of Education in Prussia and Bavaria 1750–1900
(Providence: Berg Publishers).
SCHOFIELD, R. (1973), “Dimensions of Illiteracy 1750–1850”, Exploration in Economic History, 10, 437–454.
SCOTT, F. D. (1977) Sweden: The Nation’s History (Minneapolis: University of Minnesota Press).
SIMON, B. (1987), “Systematization and Segmentation in Education: the Case of England”, in D. K. Muller, F. Ringer
and B. Simon (eds.) The Rise of the Modern Educational System (Cambridge: Cambridge University Press.)
SMELSER, N. (1991) Social Paralysis and Social Change: British Working-Class Education in the Nineteenth Century
(Berkeley: University of California Press) 88–110.
SPREE, R. (1977) Die Wachstumszyklen der deutschen Wirtschaft von 1840 bis 1880 (Berlin: Dunker & Humboldt).
TAYLOR, A. M. (1999), “Sources of Convergence in the Late Nineteenth Century”, European Economic Review, 43,
1621–1645.
THOMPSON, E. P. (1968) The Making of the English Working Class (London: Penguin).
TIPTON, F. B. (2003) A History of Modern Germany Since 1815 (Berkeley: University of California Press).
TOMES, N. (1981) “The Family, Inheritance and the Intergenerational Transmission of Inequality”, Journal of Political
Economy, 89, 928–958.
WIENER, M. J. (1981) English Culture and the Decline of the Industrial Spirit, 1850-1980 (Cambridge: Cambridge
University Press).
WILLIAMSON, J. G. (1985), Did British Capitalism Breed Inequality? (Boston: Allen & Unwin).
WOLTHUIS, J. (1999), “Lower Technical Education in the Netherlands 1798–1993: The Rise and Fall of a Subsystem”
(http://www.ub.rug.nl/eldoc/dis/ppsw/j.wolthuis/titlecon.pdf).